2009 Projection Checkpoint Through Games of June 25 - Pitching

Following up on Last week's post about their projections, here's a look at the pitching. I'm only going to look at player's who've pitched at least 20 innings in the majors for now.

One thing that needs to be noted is that no one knew how New Yankee Stadium would play, so we don't know what the park factor impact was to the projections. So make a mental note of that at the very least when looking at the numbers that follow. If the projections seem too optimistic, it's likely that the park factor is at least part of it. Projections are pro-rated to actual YTD playing time, so pay special attention to HRs since that has been the biggest issue in NYS.

C.C. Sabathia
Sabathia was the Yankees' big free agent target and acquisition this off-season. Has he been worth it so far?

C.C. Sabathia	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	102	99	42	39	9	21	89	3.41	3.23	12	25
2009 marcel projection	102	94	39	35	9	24	96	3.07	3.13	16	29
2009 pecota projection	102	98	43	39	10	23	89	3.43	3.38	12	25
2009 tht projection	102	94	40	37	10	21	90	3.25	3.28	14	27
2009 zips projection	102	98	38	35	10	21	93	3.07	3.23	16	29
2009 cairo projection	102	99	44	39	9	27	89	3.48	3.44	12	24
2009 average projection	102	97	41	37	9	23	91	3.29	3.28	14	27
2009 YTD	102	85	45	42	8	31	70	3.71	3.76	9	22

FIP: Fielding Independent Pitching
RSAA: Runs saved above average
RSAR: Runs saved above replacement

C.C.'s been a little worse than expected to this point. Interestingly, he's actually allowed fewer homers than any of the projections expected despite the homerrificness of NYS. The bigger issue is the BB rate is higher and the K rate is lower. He's giving up fewer hits than projected, as well. Still, he has an established track record of excellence and is(HOPEFULLY) healthy, so we should see him improve a little as the season goes on.

A.J. Burnett
It's funny with Burnett. A lot of Yankee bloggers and most analysts hated his signing at the time. I didn't hate it, I thought it was risky but could pay off. However, the reason everyone hated it was because Burnett was supposedly an injury risk. Not one analyst said that Burnett wouldn't pitch well when he could pitch, they just thought he wouldn't pitch enough (please correct me on this if I missed anyone). Now that Burnett's been healthy and pitching erratically, many of those same analysts are alluding that the reason they didn't like the Burnett signing was because they knew he would be erratic, not because of the health risk. I call BS.

A.J. Burnett	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	87	83	41	38	8	33	85	3.88	3.61	6	17
2009 marcel projection	87	81	42	39	9	33	84	3.99	3.72	5	16
2009 pecota projection	87	83	40	37	8	32	79	3.82	3.73	7	17
2009 tht projection	87	81	40	37	9	32	81	3.85	3.76	6	17
2009 zips projection	87	84	41	38	10	32	83	3.97	3.86	5	16
2009 cairo projection	87	83	43	39	9	26	82	4.07	3.58	4	15
2009 average projection	87	82	41	38	9	31	82	3.93	3.71	5	16
2009 YTD	87	81	44	41	13	44	82	4.24	4.77	3	13

Burnett hasn't been particularly good so far, due to giving up four more HRs and 13 more BBs than his average projection expected over the 87 innings he's pitched. While his ERA of 4.24 is respectable, it belies a FIP of 4.77 which is a non-trivial concern going forward. While it's still too early to judge his contract only 8% of the way through it, the nay-sayers appear to have been right so far. I still think(hope) A.J. can turn it around.

Andy Pettitte
Pettitte's been one pitcher who's really struggled in NYS (Opponents are hitting .332/.392/.500 against him at home, compared to .207/.277/.329 on the road). Interestingly, his combined stats are very much in-line with how he was projected.

Andy Pettitte	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	87	96	45	42	8	28	65	4.31	3.84	2	13
2009 marcel projection	87	98	46	42	9	27	63	4.38	3.99	1	12
2009 pecota projection	87	96	47	42	9	27	61	4.41	4.02	1	12
2009 tht projection	87	95	44	41	9	26	60	4.27	4.04	2	13
2009 zips projection	87	97	46	43	8	27	61	4.43	3.99	1	12
2009 cairo projection	87	96	47	43	9	26	62	4.45	3.98	1	11
2009 average projection	87	96	46	42	8	27	62	4.37	3.98	1	12
2009 YTD	87	96	44	41	11	33	57	4.26	4.68	2	13

Pettitte's not pretty to watch, but despite his ugly peripherals he's a touch above average in terms of run prevention and gives the team innings. It would be nice if he could get his home performance to match his road performance a little more closely though.

Joba Chamberlain

Joba Chamberlain	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	76	66	31	28	7	29	84	3.39	3.30	9	19
2009 marcel projection	76	67	28	26	5	28	78	3.07	3.14	12	21
2009 pecota projection	76	65	29	26	5	29	81	3.09	3.05	12	21
2009 tht projection	76	62	29	27	6	30	84	3.20	3.19	11	20
2009 zips projection	76	73	34	32	7	30	74	3.77	3.61	6	16
2009 cairo projection	76	72	31	29	7	26	77	3.45	3.36	9	18
2009 average projection	76	68	30	28	6	28	80	3.33	3.27	10	19
2009 YTD	76	70	36	32	8	37	69	3.80	4.22	6	15

Despite the Yankees' foolish insistence on using a reliever in the starting rotation, Joba's been the Yankees' second most valuable starter. That his season line seems disappointing is more a testament to how well he has done the past two seasons than any real disappointment in his current performance. He's walking a few more batters and striking out a few less batters than expected, but part of that is the fact that his projections include his relief stats, which are going to boost his projections slightly since relieving is easier than starting. As long as we remember that he's 23 and still working his way up to building the stamina to become a full-time starter, we should be encouraged by his performance to date. He's already pitched 76 innings, something that would take him a whole season to do as a reliever.

Phil Hughes
I think some people forget that Hughes is nine months younger than Joba. While his MLB time has been mixed, he's looked pretty good this year. I was at the game versus the Braves on Tuesday and the Braves fans were oohing and aahing when Hughes hit 96 on his fastball several times. The myth of his 'only 91 mph fastball' should be silenced now. A healthy Hughes has the stuff that made him one of the top pitching prospects in baseball. Now he's hitting 96 in relief, which means he's probably going to lose some of that velocity when he moves back to the rotation, but he should still maintain enough to be effective.

Phil Hughes	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	45	46	23	22	5	17	38	4.28	4.10	1	7
2009 marcel projection	45	45	25	23	5	17	35	4.61	4.19	-1	5
2009 pecota projection	45	47	26	24	5	18	36	4.74	4.27	-1	4
2009 tht projection	45	44	23	22	5	17	35	4.28	4.27	1	7
2009 zips projection	45	47	23	21	3	18	32	4.27	3.83	1	7
2009 cairo projection	45	47	24	23	4	15	34	4.53	3.94	0	5
2009 average projection	45	46	24	22	5	17	35	4.45	4.10	0	6
2009 YTD	45	42	24	23	7	17	45	4.57	4.35	0	5

As you can see Hughes is pitching around where his projections expected, although with a higher K rate. In general, a starter moved to relief will see his K rate increase by about 16%, so that explains part of the higher K rate. Still, Hughes is finally flashing the talent that made him famous. I'd still like to see him starting again, either in the majors or in AAA, because I think he needs to build his innings up.

Chien-Ming Wang
Hide the women and the kids, because what you are about to see is explicit.

Chien-Ming Wang	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	31	37	17	16	2	10	17	4.47	3.86	0	4
2009 marcel projection	31	31	14	13	2	10	17	3.83	3.89	2	6
2009 pecota projection	31	35	17	15	3	10	17	4.39	4.19	0	4
2009 tht projection	31	34	15	13	2	10	15	3.88	4.07	2	6
2009 zips projection	31	33	15	14	1	10	15	3.92	3.80	2	6
2009 cairo projection	31	33	15	14	2	11	15	4.16	4.04	1	5
2009 average projection	31	34	15	14	2	10	16	4.11	3.98	1	5
2009 YTD	31	56	40	39	6	15	25	11.21	5.53	-23	-19

If Wang can pitch 120 more innings at his average projected ERA of 4.11, he'll end the season with 151 innings and a 5.89 ERA. In the entire history of the Yankees, there have only been three pitchers to pitch at least 140 innings with an ERA of 5.5 or higher:

David Cone in 2000 (6.91 ERA in 155 IP).
Jeff Weaver in 2003 (5.99 ERA in 159.1 IP).
Kenny Rogers in 1997 (5.65 ERA in 145 IP).

The good news is Wang has been a little better lately and may be able to pitch closer to his projections over the rest of the season. Also, this may make Wang cheaper in arbitration next year. The bad news is there's very little chance of Wang salvaging this season and finishing with respectable numbers, although that doesn't really matter going forward.

Phil Coke
Coke impressed in 2008, but prior to that his minor league numbers werent' all that good, as reflected in his projections.

Phil Coke	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	30	30	15	14	3	12	28	4.25	3.96	1	5
2009 marcel projection	30	28	14	13	3	10	24	3.94	3.89	2	6
2009 pecota projection	30	34	18	17	3	12	22	4.99	4.43	-2	2
2009 tht projection	30	34	20	18	4	15	19	5.44	5.33	-3	1
2009 zips projection	30	34	18	17	4	13	19	4.91	4.72	-1	2
2009 cairo projection	30	39	21	20	4	6	18	5.82	4.33	-4	-1
2009 average projection	30	33	18	16	4	11	21	4.89	4.44	-1	2
2009 YTD	30	20	15	12	5	12	22	3.56	5.08	3	7

Now a lot of the input into his projections were based on Coke being a starter. If you convert his average projection to a relief equivalent, you get a revised line of:

IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
30	29	14	13	3	12	25	3.89	4.20	2.0	5.0

The starter->relief conversion that I use assumes:
H: 0.87
R/ER: 0.79
HR: .95
BB: 1.06
K: 1.16

So what those mean is, for example, if a pitcher is projected to give up 100 hits as a starter, he should give 0.87 times 100 = 87 hits as a reliever.

He's given up fewer hits than the relief conversion would have expected, and more HRs, but the BB rate and K rate are reasonably close. While his ERA is good, his peripherals are a warning that he hasn't pitched as well as the ERA shows. I have no idea whether we should expect his peripherals to improve to match his ERA, or if we should expect him to pitch closer to his FIP going forward.

Alfredo Aceves
In a season where almost every member of the bullpen has disappointed, AA is an oasis in a desert of suck. Aceves was projected as a starter, where he was expected to be slightly worse than average, but solidly above replacement level.

Alfredo Aceves	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	31	35	18	17	5	11	20	4.95	4.87	-2	2
2009 marcel projection	31	30	14	13	3	11	22	3.86	4.34	2	6
2009 pecota projection	31	35	19	17	4	10	19	5.03	4.84	-2	2
2009 tht projection	31	35	18	17	5	11	20	4.95	4.87	-2	2
2009 zips projection	31	36	19	18	5	8	16	5.19	5.18	-2	1
2009 cairo projection	31	32	18	17	4	4	17	4.83	4.37	-1	3
2009 average projection	31	34	18	17	4	9	19	4.80	4.75	-1	3
2009 YTD	31	25	8	8	4	6	27	2.32	3.72	8	11

Same deal as with Coke, if we convert his projection to a relief equivalent, we get an improved line of:

IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
31	30	14	13	3	12	25	3.89	4.20	2.1	5.2

FIP expects him to pitch closer to an ERA in the mid 3s going forward, looking at both his YTD performance and his revised projection, but he should still be a solid asset in the pen going forward.

Jose Veras

jose veras	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	26	24	12	12	3	11	26	4.05	3.80	1	4
2009 marcel projection	26	24	12	12	3	11	22	4.09	4.22	1	4
2009 pecota projection	26	23	12	11	2	11	25	3.70	3.75	2	5
2009 tht projection	26	23	12	11	2	11	26	3.85	3.70	2	5
2009 zips projection	26	25	13	12	3	12	25	4.31	4.36	1	4
2009 cairo projection	26	27	13	13	3	11	22	4.53	4.34	0	3
2009 average projection	26	24	12	12	3	11	24	4.09	4.03	1	4
2009 YTD	26	23	17	17	5	14	18	5.95	5.96	-4	-1

While Veras had a good season last year, he slumped later in the year, and his command was always a major concern. Veras's command was horrible this year and it's reflected in his results. He's been traded to Cleveland for cash considerations.

Jonathan Albaladejo
Albaladejo was expected to be a decent back-end relief option although his projections weren't really all that good.

jonathan albaladejo	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	21	23	12	11	3	8	16	4.75	4.67	-1	2
2009 marcel projection	21	20	10	10	2	8	17	4.09	3.90	1	4
2009 pecota projection	21	23	12	11	3	8	15	4.58	4.49	0	2
2009 tht projection	21	20	10	9	3	9	16	4.03	4.44	1	4
2009 zips projection	21	23	12	11	3	8	15	4.87	4.87	-1	2
2009 cairo projection	21	24	13	12	3	6	14	5.29	4.67	-2	1
2009 average projection	21	22	12	11	3	8	15	4.60	4.51	0	2
2009 YTD	21	24	16	14	5	10	12	6.00	6.58	-4	-1

He essentially matched Veras's output though, with a few less walks and Ks. He's doing well back in Scranton/WB and may end up getting another chance this year.

Mariano F'ing Rivera
Mo's early season velocity wasn't very good as he was building up arm strength after off-season surgery. Seeing his cutter at 88-89 mph was alarming, as was seeing batters hit .306/.306/.551 against him over his first 12 games. His velocity's been better lately and over his last 16 games he's looked more like Mo, with a 2.55 ERA and with hitters hitting .185/.221/.292 against him.

Mariano Rivera	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	29	26	10	9	2	5	28	2.86	2.64	5	9
2009 marcel projection	29	26	10	10	2	6	26	3.09	3.08	5	8
2009 pecota projection	29	25	9	8	2	5	28	2.42	2.54	7	10
2009 tht projection	29	24	9	8	2	5	28	2.52	2.57	6	10
2009 zips projection	29	24	8	7	1	4	28	2.28	2.24	7	11
2009 cairo projection	29	23	8	8	2	5	27	2.38	2.61	7	10
2009 average projection	29	25	9	8	2	5	27	2.59	2.61	6	10
2009 YTD	29	27	11	10	5	3	37	3.10	3.20	5	8

A lot has been made of Mo's high K rate, and his K/9 of 11.5 is 37% higher than his projected 8.4. Even if you instead look at the more accurate Ks per batters faced, you can see that Mo has fanned 31.6% of the batters he's faced, compared to his projected 23%. That's 37% higher. So we should be comfortable that Mo will be fine going forward, although there will almost certainly be at least one more WWWMW along the way.

Team
Here are the cumulative totals for the players above:

Total	IP	H	R	ER	HR	BB	SO	ERA	FIP	RSAA	RSAR
2009 chone projection	565	565	267	247	54	185	495	3.93	3.68	36.0	106.6
2009 marcel projection	565	544	256	236	51	187	483	3.75	3.66	46.9	117.5
2009 pecota projection	565	563	270	247	54	187	472	3.93	3.76	35.9	106.6
2009 tht projection	565	545	260	241	56	188	474	3.84	3.81	41.5	112.1
2009 zips projection	565	574	268	248	56	184	462	3.95	3.82	34.4	105.0
2009 cairo projection	565	574	278	257	56	163	456	4.10	3.75	25.4	96.0
2009 average projection	565	561	267	246	55	182	474	3.92	3.75	36.7	107.3
2009 YTD	565	549	300	279	77	222	464	4.44	4.51	3.5	74.1

The staff as comprised above has allowed 22 more HRs than expected, while walking 40 more batters and fanning 10 fewer. They've been close to three wins worse than projected to this point. At 39-32 they're on a projected 89 win pace. If they had those 3 wins, they'd be 42-29 and on a 96 win pace.

Projecting pitching is a pain in the ass, because even if you nail the peripherals, ERA is subject to fluctuations that we can't predict. If we look at the RMSE(root mean square error) for each individual pitchers' ERA, the projection systems rank like this:

Projection	rERA
chone	0.78
cairo	0.80
average	0.81
pecota	0.81
marcel	0.83
zips	0.83
tht	0.86

If we want to instead look at how well the systems did in projecting the peripherals, we can look at the RMSE of FIP, and that looks like this:

Projection	rFIP
cairo	0.83
pecota	0.84
average	0.85
tht	0.85
marcel	0.85
zips	0.86
chone	0.87

And lastly, if we average ERA and FIP and then look at the RMSE of the average, we get this:

Projection	r((ERA+FIP)/2)
cairo	0.81
pecota	0.83
chone	0.83
average	0.83
marcel	0.84
zips	0.85
tht	0.86

We do have to consider the impact of NYS on these projections, and we do also have to remember there's still close to 60% of the season left to play and a lot can change. On the whole, we can see the pitching staff has been disappointing overall, but they have been better in June and can hopefully carry that going forward.