WHIPlash: Projecting Cliff Lee’s Second Half WHIP

Yesterday, I analyzed Cliff Lee’s prospective ERA over/under projection for the rest of the season. Now, it’s time to take a look at Cliff Lee’s prospective WHIP.

In my ERA analysis, I pegged Cliff Lee for 105 innings for the rest of the season (15 starts with an average of 7.0 innings per outing). Further, in part 4 of my ERA analysis, I postulated that Cliff Lee would likely have a second half BB/9 closer to his 3 year average of 1.41 than his current rate of 0.52. Why? Well, the only pitcher I can recall to have every posted a BB/9 that low is Carlos Silva (circa 2005) and he didn’t even pitch 200 innings. Greg Maddux twice posted BB/9’s below one (0.99 in 1995 and 0.77 in 1997), but Maddux was consistently the king of control and did not even walk 1000 batters for his career. Cliff Lee has great control, but he’s not Greg Maddux. Hence, I’m going to use his 3-year 1.41 BB/9, which would give him an end of season BB/9 of 0.97.

All that is left to do in order to project Cliff Lee’s second half WHIP is to calculate his xHits total. To do this, I am going to use his first half BIP data and normalize the LD% at 19%, which is what, over the large enough sample size, pitchers tend to regress towards (Derek Carty of THT will be releasing an article about this sometime soon, I believe).

Through the first half of 2010, Cliff Lee has struck out, surrendered a home run to or walked 100 of the 408 total batters faced, putting 306 balls in play (if you did the math, you’ll notice that 408-100 is actually 308. I have no clue what happened to these other 2 mystery batters as Cliff Lee has 0 IBBs and 0 HBPs so far this season. Maybe foul outs?). Of the balls put into play, 127 of them have been groundballs (41.5%), 53 have been line drives (17.3%) and the remaining 126 balls in play have been of the flyball variety (41.2%). Breaking down the flyballs further, 115 of the 126 flyballs have been outfield flyballs, while the other 11 have been infield flyballs (8.7% IFFB%).

According to Tom Tango’s research, BABIP by batted ball type is general broken down as follows:

  • Popups: .020
  • Groundballs: 0.241
  • Outfield Flyballs: 0.177
  • Line Drives: 0.728

First, let’s normalize Cliff Lee’s LD%. At 19% of BIP, that would give Lee ~58 line drives per 306 BIP. This leaves 248 BIP to work with.

Lee has a GB/(GB+FB) rate of 50.2% and a FB/(GB+FB) rate of 49.8% for a GB/FB ratio of 1.008. Given the relative symmetry between Cliff Lee’s GB and FB rates, I am going to apportion the remaining 248 BIP evenly between grounders (124) and flyballs (124). Breaking the flyball data down further, using Cliff Lee’s 8.7% IFFB ratio, we give Cliff Lee 11 popups to 113 outfield flyballs.

Using these normalized BIP numbers (124 GB, 58 LD, 113 OFFB, 11 IFFB) and the above BABIP rates, we get a “hits allowed” total of 92.33. This, paired with his 1.41 BB/9 rate, over 105 IP, would yield a base runners allowed total of 107.78.

All in all, the above calculates into a WHIP of 1.04.

But wait, that’s not the end of our analysis. What about the Rangers’ ballpark effects and defense? Those two factors clearly have an impact on Cliff Lee’s pitching line.

Going back to our xHits formula, we need to account for the park’s dimension effects on BIP types. Hence, our xHits calculation will look as thus, based on THT’s top secret four-year park factor data: (124*.241*0.99490209)+(.177*113*0.99848143)+(.728*58*1.05542638)+(.020*11*0.96431953).

This yields a pre-defensive adjustment xHits total of 94.48 for a WHIP of 1.06 over 105 IP.

Defense is a hard adjustment to accord for. Does good defense prevent singles? triples? doubles? homers (in the case of Dewayne Wise)? Tom Tango has previously broken down runs-expectancy based on types of hits and using our previous UZR/INN rate of +0.0269 for the Rangers (see my earlier Cliff Lee ERA analysis) and 2.82 runs prevent per 105 IP, we can reverse engineer a hits prevented total. However, without accurate 1B/2B/3B/HR defensive prevention ratios, this defensive adjustment will not be perfect.

Thus, similar to how I adjust xBABIP, I am going to optimistically assume that all prevented runs would come exclusively from singles and reverse engineer a hit-prevented total based on that. As a single is worth 0.47 runs, 2.82 runs prevented would equate to 6 hits prevented. Subtracting this from our park-adjusted 94.48 xHits total, we get a ultimate xHits total of 88.48 for WHIP of 1.00 on the button.

When we take this analysis combined with my prior ERA analysis, we find that Cliff Lee will likely post an ERA between 3.13 and 3.70 with a WHIP of about 1.00 for the Rangers.

Draw what conclusions you will. As much as I like Smoak, he far from an elite hitter (Matt LaPorta, traded for other ex-Indians rental C.C. Sabathia, has more wOBA upside) and a fair price to pay for someone who will undoubtedly help the Rangers reach the post season and net them two top picks as a type A free agent this offseason. Especially when you have Chris Davis to spare.

Jeffrey Gross is an attorney who periodically moonlights as a (fantasy) baseball analyst. He also responsibly enjoys tasty adult beverages. You can read about those adventures at his blog and/or follow him on Twitter @saBEERmetrics.
Newest Most Voted
Inline Feedbacks
View all comments
Charles Saeger
12 years ago

A “penultimate” xHits total implies that there is one more step:

* “Ultimus” is Latin for “last, end”
* “Penultimus” is “paene,” meaning “almost,” prefixed to “ultimus”—“next to last” (like “paene” + “insula” = “peninsula”—“almost an island”)
* “Antepenultimus” is with what you’re probably confusing this, and thats “ante” (before) and “penultimus”—“before next to last”

Pedantic, to be sure, but I expected another step and actually had to look at everything to realize that I didn’t miss something.

Jeffrey Gross
12 years ago

Thanks for the grammar correction. I’ve eliminated the “pen” from the statement. True story, however: I used penultimate on the second to last step but thought I was misusing the term then whilst editing. Oops.