Translating Cliff Lee The Mariner Into Cliff Lee The Ranger (Parts 3 and 4)

Earlier today, in the wake of the Cliff Lee trade, I took a look at what Cliff Lee’s 2010 season would look like as a Ranger. I concluded his first half ERA would be slightly higher, at 2.49 (versus his current 2.34 mark), while his second half would still be a solid 3.13 all whilst noting that his BB/9 seemed obscenely low and somewhat unsustainable going forward. Thus, rather than look merely at what Cliff Lee might do for the rest of 2010 based on his first half numbers, I think it would be much more interesting to see how Cliff Lee would pitch based on his career peripherals or his current peripherals adjusted to his career walk rate.

First, let’s quickly recap the relevant adjustment stats:

o Park Factor Adjustment (PFA): According to Baseball-Reference.com, the Rangers Ballpark in Arlington inflates run scoring by 8%. As the Rangers only play half of their games at home, I am going to use a park factor adjustment of +4%.

o Defensive Adjustment (UZR/INN): The Rangers have a team UZR (through 07/09/2010) of +20.5 over 763.1 innings of defensive play. That equates to a UZR/INN adjustment of about +0.0269. Over 105 IP, the Rangers’ defense would prevent about 2.82 runs compared to the “league average” defensive posture.

o Innings Pitched (IP): In part two of my earlier Translating Cliff Lee post, I pegged Cliff Lee with 105 IP for the rest of the season, or 15 more starts at an average of 7.0 IP per start. I’m going to stick with this number.

And now, on to the show.

PART III: What Kind Of ERA Can We Expect For Cliff Lee Pitching For The Rangers After The All-Star Break Using His Three Year Peripherals?

For his career, Cliff Lee (the pitcher, not the early 1900′s outfielder) has an xFIP of 4.24 and FIP of 3.84. Clearly, Lee’s xFIP has been inflated heavily by early career struggles with the Indians and a look at his 2002-2007 numbers reveals almost an entirely different pitcher than his 2008-2010 numbers. Especially given the way he is pitching in 2010, using Lee’s three year peripheral data seems more relevant and indicative of his true talent level.

Over the past 3 calendar years, Cliff Lee has posted a 7.09 K/9, 1.41 BB/9 and 43.2% GB% , which is good for a 3.61 xFIP (top 20 amongst all major league pitchers who have logged at least 200 IP in that time span). Using this 3.61 xFIP as the context neutral baseline, Cliff Lee would allow 42.12 runs (RA) over 105 IP.

First, adjusting this figure for park factors, we multiply Lee’s context neutral RA by 1.04, yielding a park-adjusted RA of 43.80. Next, to account for the quality of the Rangers’ defense, we need to subtract 2.82 runs from the park-adjusted RA. Doing so, we get a park/defense adjusted RA of 40.98.

The above results in a three-year-peripheral-based 2010 second half ERA projection of 3.51.

This figure is more than a full run above the 2.34 ERA he’s sported in Seattle thus far, but it is still a plus-quality over/under ERA and worth quite a bit. Personally, I think Justin Smoak‘s projected future career is a bit overblown (I’ve compared him to Derrek Lee a la 2007 countless times), so I do not particularly think the Rangers “overpaid” to rent Lee, who will also net Texas compensation picks if offered arbitration (Lee will likely reject an arbitration offer in favor of a lucrative multi-year contract). Still, even Lee’s three year peripheral data does not satisfactorily capture his 2010 performance. This data pegs Lee with a low-7 K/9, despite the fact that his swinging strike rate is up over a full point (a 13.75% change) compared to 2008 and 2009, and a 43.2% GB%, which is 2% points higher than it is now or was last season.

PART IV: What Kind Of ERA Can We Expect For Cliff Lee Pitching For The Rangers After The All-Star Break Using His 2010 Peripherals And Three Year Walk Rate?

Enter we now part four of our analysis. Herein, we shall analyze Cliff Lee using the following to construct his 2010 second half ERA projection: his 2010 K/9, his 2010 GB% (from which we will calculate xHR/9) and his three-year average BB/9. The first and third rates are already pre-calculated at 7.73 and 1.41, respectively.

What remains to be calculated is the xHR/9 rate from which we can derive his “xxFIP” (expected xFIP, a stat I completely made up to represent adjusted xFIP). To calculate Cliff Lee’s xHR/9 rate, we need to do two things. First, we need to determine his context neutral home run rate based on an 11% HR/FB%. Second, we will need to account for the Rangers’ 1.10025793 HR/FB% park factor (per THT’s top secret four year park factor data). As the Rangers only play half their games at home, the true multiplier is, of course, 1.050129.

To calculate Lee’s context neutral home run rate based on an 11% HR/FB%, I am going to use his first half BIP data. Through his first 103.1 IP, Cliff Lee has surrendered 126 flyballs (and only 5 HR, for a ridiculously low 4.0% HR/FB% which is low even by his 8.1% career mark). With an 11% HR/FB rate, Cliff Lee, in theory, should have given up 13.86 HR for a neutral 1.21 xHR/9 rate. Adjusting for the Ballpark in Arlington’s HR/FB% park factor, this xHR/9 rate inflates to 1.27 HR/9 (14.55 HR surrendered for 103.1 IP). This 1.27 figure is what we will use to calculate Lee’s prospective “xxFIP.”

The formula to calculate FIP (or xFIP in this case, as we are using an xHR/9 rate) is 3.2+((13*HR)+(3*(BB+HBP-IBB))-(2*K))/IP. Using the 105 IP projection for Cliff Lee’s second half, that gives Lee 14.82 HR, 16.45 BB and 90.18 K. Given the relatively similar sample sizes between Cliff Lee’s actual 2010 innings total to date to his projected IP (103.1 versus 105) and his pin-point control, I am going to hold his current HBP-IBB amount constant at zero.

Using the above, we find that Cliff Lee’s 2010 second half “xxFIP” would be 3.79. Over 105 IP, this “xxFIP” baseline would give Cliff Lee 44.18 runs allowed (RA). Adjusting for both park factors (*1.04) and defense (-2.82), Cliff Lee’s prospective second half RA total, based on his “xxFIP,” would be 43.13.

Over 105 IP, this means a 3.70 ERA. Now a 3.79 figure may seem to be the most pessimistic of any of the baselines from which I projected Cliff Lee’s prospective second half over/under ERA on the Rangers, but it makes the most sense, at least to me, when you consider how obscenely low his current HR/FB and BB/9 rates are.

However, if you believe that Cliff Lee’s career HR/FB rate is better indicative of the rate to which he should “regress” to, we should do the above analysis with an 8.5% HR/FB rate (a park adjusted figure based on his career 8.1% mark). Using this 8.5% mark, Cliff Lee’s “xxFIP” would be 3.37 and his adjusted second half ERA over/under projection would be 3.26.

CONCLUSION

As I pointed in part 2 of this analysis, Cliff Lee is a very good pitcher. My above calculations peg his second-half over/under ERA projection for 2010 somewhere between 3.13 and 3.70. Obviously, the Rangers hope Lee posts an ERA below 3.7, but if that’s his prospective downside, well, it makes a lot of sense for a playoff contender to pay big for that kind of production. From a fantasy perspective, Lee only becomes more valuable. No right-headed owner expected Lee to continue posting a sub-3.00 ERA and the Rangers quality defense somewhat offsets the (overblown) change in scenery. If you can find an owner willing to sell at a slight discount due to a worry about how Lee will pitch mid-summer in the Ballpark at Arlington, you should jump on that opportunity immediately.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Translating Cliff Lee The Mariner Into Cliff Lee The Ranger (Parts 1 and 2)
Next: Does every *seller* get one representative? »

Comments

  1. kds said...

    1)  Haven’t you estimated RA, not ERA?  So the ERA would probably be lower.

    2)  Has anyone looked at the relation between park factors and IP/GS?  Component park factors would be best, but even with just the runs adjustment one might think that a Ballpark like the one in Arlington with its increase in scoring would lead to an increase in pitch count and a decrease in IP.

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>