#1, #2, #3, #4, and #5 NL Starters At The Halfway Point

by James Kannengieser on Jul 7, 2009 7:30 AM EDT in Analysis Comment 14 comments

When describing starting pitchers, we often say things like "he's an ace" or "he's a #4 pitcher". A pet peeve of mine is the wrongful labeling of a pitcher to a certain rotation spot, based on subjective perception rather than objective fact. Poll a group of baseball fans about this topic and you will likely receive a vast array of opinions. My take, which is similar to the ones shared in this post and this post, is that if a pitcher is in the top 1/5 of starters in his league, then he should be called a #1 starter. The next 1/5 are #2's, the next 1/5 are #3's, etc. I pulled some statistics from Fangraphs to devise a quick and dirty way of classifying National League starters at the halfway point of the season (the Mets have played 81 games, hence I'm calling this the halfway point).

A total of 7,613 innings were thrown by 121 different NL starters through Sunday. I put together a spreadsheet of all these pitchers, their innings pitched, and their FIP (fielding independent pitching). Dividing 7,613 by 5, I created 5 groups of 1522.2 innings. Beginning with the pitcher who boasts the best FIP (P.J. Walters, 1.04 FIP through 4 innings pitched), I summed innings until reaching 1522.2 to determine #1 pitchers. I did the same all the way through the bottom spot on the list, filled by John Koronka and his 11.90 FIP. Standard caveats apply here - FIP doesn't account for batted balls like tRA, it isn't park adjusted, etc., but it's better than ERA for evaluating a pitcher's performance. Here are the results, showing pitcher classification, the number of pitchers that fit that classification, and the corresponding FIP:

Rank	Pitchers	FIP Range
#1	22	0.00-3.53
#2	19	3.54-3.95
#3	19	3.96-4.55
#4	24	4.56-5.09
#5	37	5.10+

Keep in mind that of the top 22 pitchers, only 9 have thrown atleast 100 innings. Rodrigo Lopez owned the Mets on Friday night and has a 2.25 FIP through 6.1 innings, but calling him a true #1 is a stretch. Here is how the Mets' pitching staff stacks up, using this method:

Pitcher	IP	FIP	Rank
Jon Niese	10.2	1.72	High #1
Johan Santana	109.1	3.91	Low #2
Mike Pelfrey	88.2	4.10	High #3
Livan Hernandez	96.2	4.68	High #4
Fernando Nieve	27.2	4.74	High #4
John Maine	61.2	4.90	Low #4
Tim Redding	47.2	5.05	Low #4
Nelson Figueroa	6	5.70	Low #5
Oliver Perez	21.2	6.04	Low #5

While it's true that Jon Niese should be freed from his minor league shackles, it's doubtful he would mantain his high #1 status. This means the Mets' best starter has been the equivalent of a low #2 pitcher, and most innings pitched by starters have been by #4 and #5 types. Only 14 starters have been worse than Ollie, but he'll get a shot to improve on that starting Wednesday night. Peruse the spreadsheet if you like, and here are some notable names and their classifications:

Carlos Zambrano
Roy Oswalt
Johnny Cueto (maybe not after last night's Phillies shellaquing)
Randy Wolf
Ryan Dempster

1 recs | Comment 14 comments | Share

Story-email Email Printer Print

Comments

Nice work James

We have a pretty crappy rotation

And doesn’t it seem like Ollie pitched a lot more than the 21.2 excruciating innings he pitched?

"We must win and we must know how to win rather than win because we have statistical people."

by Evan_S on Jul 7, 2009 7:41 AM EDT reply actions 0 recs

cool post

however i do think a healthy amount of subjectivity is appropriate for determining a pitcher’s “number” in the rotation. i heartily agree that people overestimate the ability of a pitcher’s supposed number and think most also overestimate the number of pitchers that have the requisite capabilities to be labeled in each of these rotational slots. but i think some of that is good. i dont think there should be 30 “# 1” picthers int he majors. i think its more like between one half and 2/3 of all teasm ahve a legit “#1”. theres probably a little less than 30 #’2, #3 should be about average pitcher and a 4 fill sin the gap between a 5, which woul dbe replacement picther, meaning every team as a nearly unlimited supploy of #5 pitchers, though not all teams actually start a #5, like if your rotation is particularly stacked.

of course, such lack of objective standards means the bloviating scotties and fancesa’s of the world will undoubtedly louse up the whole concept so James’ approach is certainly appreciated. with that in mind my off the cuff, unresearched (except with some of the stats in the post above) labeling of the mets rotation would be

santana 1 (this is a no doi)
pelfrey 2 (this is debateable, if you want to call him a 3 i wouldnt argue)
maine 3 (based on prior preformance)
fatty fat fat livan 4 (based on this year’s preformance – notice my blatent lack of consistency)
everyone else 5
ollie 37 see you in hell oliver

Lets hope that when gut check time comes again the Mets will pass it with flying colors.

by kendynamo on Jul 7, 2009 8:31 AM EDT reply actions 1 recs

Question about the method

How come you didn’t use the sortable FIP stats from the hardball time and limit results to qualified pitchers?

by Reg Dunlop on Jul 7, 2009 8:38 AM EDT reply actions 0 recs

I wanted to factor every inning pitched by an NL starter

so I didn’t limit it to qualified pitchers.

by James Kannengieser on Jul 7, 2009 10:04 AM EDT up reply actions 0 recs

While this method

is helpful in getting some perspective in how our rotation has fared this year, I’m not sure it’s a good way to label pitchers. It’s a little too fluid – as you pointed out, Cueto was a #3, for his next start he’ll be a #4, maybe if he pitches well he’ll be back to a #3. And if you only consider a #1 an “ace,” that can fluctuate from start to start. Johan’s an ace! (Yankee game) No he’s not! (Perfect game against the Reds on Sat.) Now he’s an ace! It can vary too much to be useful. And if someone pitched in the top fifth one year, and is now pitching in the second fifth, I would consider him an ace (and so on).

by Mount17 on Jul 7, 2009 9:37 AM EDT reply actions 0 recs

It's not a method for definitively labeling pitchers

It’s just a way to see how pitchers have fared through a half season’s worth of innings. By no means am I saying “Ricky Nolasco is a #1 pitcher and will be forever and ever, Amen.” For the first half of the season, this is how pitchers have performed.

by James Kannengieser on Jul 7, 2009 10:10 AM EDT up reply actions 0 recs

Sure thing

by Mount17 on Jul 7, 2009 10:15 AM EDT up reply actions 0 recs

Really cool post, James

by Pat Andriola on Jul 7, 2009 10:05 AM EDT reply actions 0 recs

I keep thinking about Sam's comments

about using the result to determine the process ad-hoc. It’s certainly fair to worry about Johan’s mechanics and peripherals but we shouldn’t let a couple of 2009 outliers define his rank as a pitcher.

However, though it’s too soon to use these FIP rankings to define our rotation for the 2009 season, they do succeed in confirming this season’s story so far: Johan has been great but a disappointment, Ollie has blown the big one, Livan has been mediocre but better than expected, everyone else has been pedestrian, and Jon Niese should be in the majors.

by TheBigStapler on Jul 7, 2009 10:14 AM EDT reply actions 0 recs

I'm not so worried about Johan.

I fully expect him to enter his second half domination run any day now.

"I dunno. I never smoked any Astroturf"
-Tug McGraw

by squid92 on Jul 7, 2009 11:55 AM EDT reply actions 0 recs

Remember the Maine

For unknown reasons, I still believe in Maine. I’m always surprised when he does not pitch well. When he hits his spots, he’s a really good pitcher in my opinion. I think he’s the X factor. If he can perform well in the second half, they have a chance. If he doesn’t come back, they’re in trouble because that means another spare part is getting too many starts. Maybe he needs to be in the pen, but I don’t know if his arm could take that or not.

My ideal starting 5 in August & Sept. would be: Johan, Pelf, Maine, Niese, Ollie/Livan/Nieve (whoever looks the best out of those last 3). I’m not sold on having Livan pitch for the entire year. They need to give him a phantom injury at some point and DL him.

by David G on Jul 7, 2009 12:03 PM EDT reply actions 0 recs

Charlie Manuel...

should have used this chart to pick the All-Star staff. Maybe then Javy Vazquez would have been chosen instead of Santana. Or Jurrjens… or even D.Lowe. All have been better than Johan.

"It breaks your heart. It is designed to break your heart. The game begins in the spring, when everything else begins again, and it blossoms in the summer, filling the afternoons and evenings, and then as soon as the chill rains come, it stops and leaves you to face the fall alone." A. Bartlett Giamatti

by sddbaker on Jul 8, 2009 1:51 AM EDT reply actions 0 recs

why not using weighting?

Rather than using “summed innings until reaching 1522.2 to determine #1 pitchers”, I think it makes more sense to take a weighted average of the FIP (weighted by IP). Then, find the 20th percentile, 40th percentile, 60th percentile, and 80th percentile, and use those as the cut points. Any pitcher who is below the 20th percentile is a #1, between the 20th and 40th is a #2, etc.

I took the liberty of calculating this weighted average: 4.29, with a standard error of the mean 0.012.

The percentiles are as follows:
0th – 1.04
20th – 3.53
40th – 3.95
60th – 4.55
80th – 5.09

Looks like the results are basically the same, but I thought it was worth mentioning…

by gabefarkas on Jul 8, 2009 1:22 PM EDT reply actions 0 recs