Beyond the Box Score: An SB Nation Community

Navigation: Jump to content areas:


Sports blogs for fans, by fans.
New Blog: The Boxing Bulletin for Boxing Fans!

Cliff Lee vs. Cole Hamels: Graph of the Day

Lee_vs_hamels-late2009_medium

In his last 21 starts, (Joe) Blanton is 9-4 with a 2.78 ERA, allowing only 125 hits in 139 1/3 innings. And, yeah, that might make him the Phillies' best pitcher this season.

Only it really won't. From the Philadelphia Daily News, and only 8 days ago

Star-divide

First a disclaimer. The FIP (Fielding Independent Pitching) numbers featured above don't have the National League constant adjustment figured in. So if that ruffles you in any way, add 3.22 to the numbers you see above and wonder afterwards if it was worth the effort.

And of course, low FIP is good. High FIP is bad. So those jutting peaks of a mountainscape you see above are indicative of some pretty rough down the stretch performances. Of course, in seasonal FIP calculations, those numbers don't weigh as heavily as the good start - less innings pitched in all that. But in the case of a start by start basis, it works.

What's the most telling story that jumps out from this chart? That both Cliff Lee and Cole Hamels have had trouble in a couple of stretches since the trade deadline? That's really no surprise to anyone even remotely close to following the organization ... I'm pretty sure with just a little bit of digging we could find any number of stories decrying the end of both of them as an ace, dated within 6 weeks of each other. And probably a couple dozen cases begging Charlie Manuel to start J.A. Happ, Pedro Martinez, Joe Blanton, heck, probably Jamie Moyer as their #1 or #2. So that's been done.

Maybe some normalization as we get closer to the current date? Those numbers at the end are awfully close to their career stats. That certainly seems reasonable.

What I see (small sample asterisk firmly attached) is an odd pattern where neither player has managed to be consistently effective at the same time. And there's really no rhyme nor reason to it. Both players pitched against the same team in the same park with roughly the same lineup 7 out of the 11 times (also of note: a reason that using FIPx wasn't really necessary). There's no dramatic split issue to face. I find this sort of comforting.

The information we as fans have found wanting (the type of analysis that led us to be here, now in the first place) drastically overreacts to these sort of odd sways and shifts, constantly seeking some sort of sign that there is either doom or absolution on the horizon, every hit a sign of weakness, every strikeout one of invincibility. Things have a tendency to even out. Regression. Predictability. Comfort. It doesn't just make our hobby easier, it makes it possible. No correlation means no indication of future success, and that goes for both scouts and analysts. That means you grab 9 guys off the street and give them bats and gloves. That's a bad sport, albeit with some excellent contest prizes (fifth caller is the right fielder!).

But yet in a week it's like we play a different game. I have never in my adult lifetime been a fan of a playoff-bound baseball team. That alone carries it's own special brand of angst, I suppose, but in this particular case I am relieved. Because within the confines of a small series that will make or break the general perceptions of a team or player for decades to come, this sort of seemingly random statistical anomaly would absolutely terrify me. One only has to scan down the Phillies roster until you get to Chan Ho Park to remember the sort of brutal legacy that could be looming.

If this sort of short term meltdown happens in the playoffs for Cole Hamels, he might be safe. He still has some goodwill built up. But anyone else? You can imagine the headlines. Cliff Lee? Couldn't cut it on the biggest stage. A compiler. Joe Blanton? Too heavy. Poor conditioning made him wear down. (And seriously, isn't Joe Blanton like a year and a half overdue to give up like 9 HR in a game?) J.A. Happ? Charlie Manuel should've known better than to play a rookie in a big spot. Heaven forbid, Chan Ho Park again? I'm not sure polite society would allow the reprinting of what could spew forth. It would be lazy sportswriter heaven, and they'd pick his bones clean before morning.

And that's just one team. Pitchers (particularly the top 2 starters) bear an unbalanced portion of post-season success. And both of these guys do come packaged with the expectation of dominance, and anything less could be, if not eternally damaging, certainly nagging. More superstitious front offices could cost these guys millions.  So, as baseball fans we probably know better than to hope for cooler heads to prevail(especially, say, the Philadelphia, New York or Boston fans). But as a (hopefully) reasonable and pragmatic subset of baseball fan, it's probably a good idea to have the ice packs ready.

 

Data Visualization: Justin Bopp | Data Collection: Justin Bopp & Walter Fulbright | Data Source:Fangraphs.com

1 recs  |  Comment 15 comments |

Story-email Email Printer Print

Comments

Display:

so

what are you saying?

by jamiethekiller on Oct 2, 2009 9:07 AM EDT reply actions   0 recs

I was wondering the same… the post seems to have gotten way off track halfway through

by BSizzle on Oct 2, 2009 10:01 AM EDT up reply actions   0 recs

I dont..

see much of a point in this article either. Lee and Hamels will undoubtedly be the 1 and 2 starters for the Phillies in the post-season.

by nmigliore on Oct 2, 2009 10:14 AM EDT reply actions   0 recs

I think the point is that start-to-start FIP tracking (or other averaged stats) during the regular season

…can show ups and downs that season-long averaged stats will smooth over.

Even two guys as dominate as Lee and Hamels are prone to short-term anomalies that appear as minor blips in a season, but in the playoffs, one (or two) of those peaks occurring in the same series could bring the Phillie’s playoff run to a quick stop, something more pronounced when we’re talking about the Ace and #2.

"What we do in life, echoes in eternity!"

by Justin Bopp on Oct 2, 2009 10:34 AM EDT reply actions   0 recs

Might be best to use Game Scores in this case instead of FIPs

Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.

by Jeff Zimmerman (TucsonRoyal) on Oct 2, 2009 10:57 AM EDT up reply actions   0 recs

Why?

For quantity? Maybe then simply (RepFIP – FIP)*IP/9 for each game? “WAR per game?”

by Sky Kalkman on Oct 2, 2009 5:37 PM EDT up reply actions   0 recs

I just found the use of FIPs confusing. FIPS not normally used for individual games and didn't know if it was thier overall FIPS changing or not.

WAR per game would be bad either.

Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.

by Jeff Zimmerman (TucsonRoyal) on Oct 2, 2009 6:03 PM EDT up reply actions   0 recs

okay

so whats the deal with sports writers and headlines and blanton giving up 9 homeruns.

to me it seemed like there’s chance they both will suck in the playoffs and that philadelphia will run them out of town. thats the feel i got.

by jamiethekiller on Oct 2, 2009 11:00 AM EDT up reply actions   0 recs

Newbie Question on FIP

I’m not sure I really trust FIP as a tool to evaluate a pitcher’s performance, at least not in the short-term. A ground-ball pitcher who (for example) surrenders three doubles in a row on a regular basis is not penalized at all by this metric. Does FIP assume that the pitcher’s BABIP will regress to the mean, or is every ground ball a pitcher surrenders no longer his fault, and thus becomes “fielding independent”?

Seeing as how FIP is a fairly well-established stat, I’m sure there is a well-established answer for clueless folks like me. ;-)

Formerly known as "Andersklasen."
Check out the best Twins' blog on the web: TwinsTarget.com.

by TwinsTarget on Oct 2, 2009 6:02 PM EDT reply actions   0 recs

FIP assumes every ball in play is of equal value.

In other words, it regresses BABIP 100% and doesn’t care about the difference between GB, FB, and LD.

by Sky Kalkman on Oct 2, 2009 6:04 PM EDT up reply actions   0 recs

Hmm.

Shouldn’t it be used, then, completely for predicting the future? What’s different about xFIP? Again, sorry for my lack of knowledge on these matters, but I don’t see the merit of FIP as an evaluation tool.

Formerly known as "Andersklasen."
Check out the best Twins' blog on the web: TwinsTarget.com.

by TwinsTarget on Oct 2, 2009 6:32 PM EDT up reply actions   0 recs

It's a very good question.

xFIP substitutes FB% for HRs, assuming all pitchers give up the same number of HRs per fly ball. Over smaller samples, it’s probably more accurate than FIP in predicting the future, but pitchers certainly do control their HR/FB rate, it just takes longer than K% or BB% to stabilize (and a lot less time than BABIP).

tRA is like FIP, but uses league-average rates for different types of batted balls: GB, FB, LD mostly. Giving up LD is bad.

Now, the question is, to what degree to pitchers control LD rate? And in a given game, if a hitter swings at a ball in the dirt and hits a liner to CF, was that a bad job by the pitcher or just one of those things that happens? What if he throws an 85 mph fastball right down the middle and the hitter pops it up? Good pitch? Probably not, but what if he had set the hitter up for it? Is that possible?

I think pitching is one of those things where constantly looking at finer and finer data is both useful and not useful. It helps us understand the big picture, but might not be meaningful in itself.

by Sky Kalkman on Oct 2, 2009 7:17 PM EDT up reply actions   0 recs

If you want to know the exact value of each batted ball, the best we currently have (well, in theory) is PZR.

It’s the other half of UZR. A pitcher gives up a batted ball and you figure out how difficult it is to field given location, speed, handedness, etc. Credit/blame the pitcher for that. That’s PZR. Then take the actual result of the play and compare it to the expected value of the play and credit/blame the fielder for that difference. That’s UZR.

by Sky Kalkman on Oct 2, 2009 7:21 PM EDT up reply actions   0 recs

Thanks for the help.

For the example above, the Lee/Hamels game-by-game chart, would FIP really be the best option? Maybe PZR or game score, which include more elements, would be the better option?

Formerly known as "Andersklasen."
Check out the best Twins' blog on the web: TwinsTarget.com.

by TwinsTarget on Oct 2, 2009 7:25 PM EDT up reply actions   0 recs

Comments For This Post Are Closed


User Tools

We use numbers and stuff.
Community Guidelines
Why be a member?
Start posting on Beyond the Box Score »

Join SB Nation and dive into communities focused on all your favorite teams.

FanPosts

Community blog posts and discussion.

Recommended FanPosts

Small
PZR-based Win Values 2001-2006
Small
The "30 parks on a budget" challenge

Recent FanPosts

Small
Determining Batted Ball Rates using Pitch Type and Location
Small
a new xBABIP calculator
Img587561916661595
Top 15 high school MLB draft prospects
Sunflower_small
World Series Simulation, Game #6
Small
JT20 Dynasty League
E52205a2_small
New Look
Sth70021_small
Exploring Hit f/x, Albeit Badly
Redcap_small
Ricky Nolasco: 4 WAR or 1 WAR?

+ New FanPost All FanPosts >

FanShots

Quick hits of video, photos, quotes, chats, links and lists that you find around the web.

Recent FanShots

Primer on BaseRuns
Cool Baseball Infographics
ESPN's Jerry Crasnick on defensive metrics
I’m also a follower, since Brian Bannister’s on our team, of sabermetric st...
Top Ten Baseball-Reference.com's Sponsorships
Primer on Linear Weights
JC Bradbury on "Hot Stove Myths"
Everyone Should Learn to Throw a Cutter
Criminals of WAR
Ten statisticians you should know about

+ New FanShot All FanShots >

BtB on Twitter

Main Feed: @BtBScore

Tommy B: @tommy_bennett
Sky: @BtB_Sky
Dan: @dturkenk
Harry: @harrypav
Jinaz: @jinazreds
Jack: @jh_moore
Erik: @Erik_Manning
Tommy R: @trancel
Justin: @justinbopp

Subscribe to BtB via Email

Enter your email address:

Delivered by FeedBurner

BtB Goes Social


Managers

Nando_small R.J. Anderson

Limes_125_small Sky Kalkman

E52205a2_small Tommy Bennett

Editors

Face_small Harry Pavlidis

Rawlings_baseball_bigger_small Dan Turkenkopf

770insig_small Jeff Zimmerman (TucsonRoyal)

Aviles_small Justin Bopp

Authors

Banny_small erik

Raysring1_small Tommy Rancel

Jinaz-reds-avatar_small JinAZ

Jmlogo_small Jack Moore

1753738656_110919ebe9_o_small vivaelpujols

1_small Graham

Baseball_small Mike Rogers

Redcap_small SFiercex4

Small Patrick Clark

Walter_album_small Walter Fulbright