Comparing Livan Hernandez and Randy Johnson

Prior to the start of the 2009 season, the thought of comparing Livan Hernandez and Randy Johnson was ludicrous. Mets fans rolled their eyes when the club signed Hernandez while Giants fans were quite pleased when the team added Johnson. To support this belief, Johnson had an ADP of 141 while Hernandez did not crack the top 348 players listed over at Mock Draft Central.

Here were the fantasy stats for both players in 2008:

RJ – 11 W, 3.91 ERA, 173 K, 1.24 WHIP
LH – 13 W, 6.05 ERA, 67 K, 1.53 WHIP

To further drive home how big the gulf between these two players was last year, RotoTimes had Johnson with a $7.19 final dollar value compared to a minus $14.92 for Hernandez.

Right now each pitcher has made eight starts and here are their fantasy stats:

RJ – 3 W, 6.86 ERA, 42 K, 1.55 WHIP
LH – 3 W, 4.93 ERA, 25 K, 1.47 WHIP

RotoTimes values

RJ – minus $16.59
LH – minus $11.66

Many people believe that Johnson will turn things around while next to no one believes in Hernandez. With Johnson, they’ll point to his BABIP or his HR/FB and talk about regression or they’ll talk about other metrics and see how similar they are to last year and conclude that a return to the Johnson of old is just around the corner.

One area where there is a big difference is in pitch selection. Overall, Johnson threw 51.4 percent fastballs last year, compared to 48.8 percent this season. With his slider it is 40.4 percent this year compared to 35.2 percent in 2008. But if we look at what Johnson was doing at the end of last season, we get an even bigger difference. These numbers come from Dan Brooks’ wonderful site.

9/1	66	12	3	 	
9/13	61	8	9		
9/18	60	15	3	2	
9/23	54	10	26		5
9/28	65	11	4	1	
Pct     75	14	11	<1	<1

He was throwing his fastball nearly seven times as often as his slider. The two outings where he threw more than a handful of sliders, his average vertical break was over three. It was 3.43 on 9/13 and 3.03 on 9/23.

(Another thing that may jump out at you is that FanGraphs does not show Johnson throwing a changeup last year. My best guess is that what Brooks labels as a changeup FanGraphs categorizes as a splitter. This does not affect the fastball-slider ratio.)

Flash forward to this year and Johnson has been all over the map with the vertical break on his slider. His first three starts it was 2.89 then 5.54 then 2.84. And since then it’s been even more erratic. Here are Johnson’s starts since April 25th:

	IP	ER	BB	V-Break
4/25	3.1	2	7	1.79
5/1	7	0	0	4.16
5/6	5.2	7	3	2.16
5/11	5	4	0	2.53
5/16	4	7	0	2.22

If we look at the Pitch Type Linear Weights, we can see the deterioration of Johnson’s slider. Since an injury-shortened season in 2003, here are his yearly wSL (which stands for runs above average with his slider):

2004 – 32.5
2005 – 24.1
2006 – 14.9
2007 – 10.1 (injury shortened)
2008 – 5.2
2009 – negative 1.0

Johnson’s command was poor in April (13 BB in 19 IP) but very good in May (3 BB in 21.2 IP). His BABIP was .268 in April and .394 so far this month. Perhaps Johnson’s command and BABIP having an inverse relationship is an insignificant small sample size fluke. Or perhaps his slider is no longer a reliable pitch and he has to go back to throwing more changeups/splitters. Or maybe at age 45 he is done as a worthwhile fantasy pitcher.

One can count on the fingers of two hands the number of pitchers who have thrown 100 innings at age 45 in MLB history. Most of them are like Jamie Moyer, crafty soft tossers. The best comparison to Johnson is Nolan Ryan. Here are the fantasy stats for Ryan at age 44 and 45, when he made 27 starts both seasons.

1991 – 12 W, 2.91 ERA, 203 K, 1.006 WHIP
1992 – 5 W, 3.72 ERA, 157 K, 1.316 WHIP

All great pitchers reach the end of the road. Ryan was done after 13 starts and a career-high 4.88 ERA in 1993. If a pitcher in his prime had Johnson’s 2009 numbers after eight starts, regression would be the most likely occurrence. But perhaps this is not bad luck but rather the new level of performance for the Big Unit. Time will tell.

Now let’s look at Hernandez.

He had another fine outing last night, going seven innings and allowing just one run. The offense took the night off for him and he wound up with a no-decision in a game the Mets went on to lose. That brings the team’s record in starts by Hernandez to 5-3. He certainly pitched well enough to win last night and in another team loss, the bullpen gave up four runs and allowed an inherited runner to score that went on Hernandez’ record.

The alleged 34-year-old Hernandez, the pitcher no one wanted, has given his team a chance to win in seven of his eight games and he has four Quality Starts this season (Johnson: 2).

Hernandez claimed in Spring Training that his knee was 100 percent. This year his average fastball is up nearly a mile per hour (83.7-84.6). Additionally, Hernandez has thrown his slider more often (16.6%-21.2%) and with better results, as his wSL this season is 2.2 after being a negative 10.2 last year. This has led to a SwStr% of 5.5 in 2009 for Hernandez after it was 3.8 last year with Minnesota and 3.0 in Colorado.

With a .316 BABIP (career rate .310), 71.7 percent strand rate (career: 72.4), HR/9 of 1.18 (career: 1.05) and 4.65 FIP (career: 4.45), Hernandez has not been the recipient of a huge amount of luck. The numbers stand out in stark contrast to what he accomplished last year but the knee injury could at least partially explain 2008’s awful numbers.

Right now, Hernandez and Johnson are only worthwhile in deep NL-only leagues. Hernandez is healthy and basically pitching to his career averages in the four metrics listed above. There have been no reports of an injury to Johnson, but he is pitching far worse than was expected coming into the season. Before you repeat the regression mantra, track what he does with both his command and his slider in his next few outings.

Print This Post

6 Responses to “Comparing Livan Hernandez and Randy Johnson”

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

    I just want to point out Eric and Dave absolutely killed the DBacks for passing on Randy and signing Garland. I understand Randy had a much better season in ’08 but it’s not as if it was crazy to assume there’s a good chance Randy’s performance will drop sharply. He’s a 45 year old power pitcher ffs.

    Meanwhile, in about a quarter of the season, Garland has been worth 0.4 WAR, which would make him on pace for about 3.2 WAR during the length of his contract. If we take into account the ZIPS mid-season projection (4.5 FIP for the rest of the season), he’ll get closer to 2 WAR for the season. So that makes Garland a good investment, and Randy, not so much.

    Vote -1 Vote +1

  2. cowdisciple says:

    RJ’s xFIP is 3.63. His ERA is twice that almost entirely due to a ridiculously unlucky 31.6% HR/FB. His 65.3% strand rate is also a little on the low side. All his controllable skills are almost exactly identical to where they have been the last 2 years. Livan’s xFIP is 4.54, which suprisingly makes him a moderately useful real-life pitcher, but does not make him Randy Johnson. Strongly disagree on this one.

    Vote -1 Vote +1

  3. Brian Joura says:

    FanGraphs has Johnson with a 24.4 percent HR/FB rate.

    The updated ZiPS shows Johnson with a drastically reduced HR rate the rest of the season and still finishing with a 4.82 ERA from now through the end of the year.

    Essentially, Johnson is a two-pitch pitcher with his FB and SL. Linear Weights analysis shows that his FB has not been a plus pitch since 2004. And now, after five consecutive years of consistent decline, his SL is no longer a plus pitch, either.

    At this point in his career, Johnson has virtually no room for error. When he has his good stuff, he can win and even dominate teams. But no pitcher has his good stuff all of the time and Johnson, as unique and as great as he’s been in his career, is no exception.

    Regression is going to come (if it doesn’t he’ll retire) but if you think he’s going to have a 3.63 ERA the rest of the season, I think you’re kidding yourself. is having some problems. Right now the Giants have played 40 games but they show splits for only 38. But in those 38 games, the Giants and their opponents have hit 23 HR in 21 games at AT&T Park. In road games for the Giants, the two teams have combined for 32 HR in 17 games. That works out to a simple HR factor of .582 thru 38 games. Imagine what Johnson’s HR numbers could look like if he was still in Arizona, which has a simple HR factor of 2.009 after 38 games.

    Vote -1 Vote +1

  4. cowdisciple says:

    Actually I got that HR rate from THT. Perhaps they haven’t updated recently. Still, if Johnson is going to fall off a cliff, shouldn’t we see a collapse in his K rate, etc? If his FB and slider have both been junk since ’03, how has he pitched so well? I don’t understand the pitch linear weighting well yet — if 4-5 of the HRs came on sliders, would that account for the regression in the linear weight?

    Vote -1 Vote +1

  5. Brian Joura says:

    I think THT uses a different denominator in figuring out their metric.

    Here’s a good piece on the LW numbers:

    I’ve only seen two of his starts this year, but I think Johnson’s problem is consistency with his pitches. His good slider is still getting people out but the trouble is that there just aren’t as many good sliders these days.

    Perhaps someone who has seen more of his games can chime in with their observations.

    Vote -1 Vote +1

  6. cowdisciple says:

    Also, I don’t know that I trust ZiPS for a player so far on the tail of the age range — there just isn’t enough data on 45-yr old pitchers to really know what to expect.

    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 *