This is one way of looking at the role of the starting pitcher: Did he give his team a better than average chance of winning the game? In an attempt to answer this question every 2007 start by a NL-Central starting pitcher was broken down into one of the following four categories:
1)Quality Start (QS) - this is defined as having pitched 6 or more innings and given up 3 or less ER.
2)Bad Start (BS) - this is defined as having pitched less than 6 innings and given up more than 3 ER.
3)Quality Innings Pitched Start (QIP) - this is defined as having pitched 6 or more inning and given up 4 or more ER.
4)Quality Runs Start (QR) - this is defined as having pitched less than 6 innings and given up 3 or less ER.

ALL 381 games started by NL-Central starters were logged into one of these four categories, as well as being noted whether that start was won or lost.
The results were:
Code:
QS - 183/116 63.4%
BS - 84/15 17.9%
QIP - 61/11 18.0%
QR - 53/27 50.9%
Next a list was compiled of every pitcher who started a game in 2007 for a NL-Central team:

Code:
Pitcher	STARTS	QS	BS	QIP	QR	WS%	AGW%
Sampson	11	8	1	0	2	90.9%	57.0
Sheets	13	10	1	1	1	84.6%	55.4
Gorzela	13	9	0	2	2	84.6%	54.5
Snell	13	10	2	1	0	76.9%	52.9
Marquis	13	7	2	1	3	76.9%	50.0
Hill	13	9	1	3	0	69.2%	49.4
Looper	13	9	3	1	0	69.2%	49.4
Oswalt	15	10	1	4	0	66.7%	48.3
Vargas	11	5	3	0	3	72.7%	47.6
Arroyo	14	8	3	2	1	64.3%	46.3
Lohse	13	8	3	2	0	61.5%	45.9
Lilly	13	8	4	1	0	61.5%	45.9
WandyR	12	5	3	1	3	66.7%	45.1
Harang	14	8	4	2	0	57.1%	43.9
Wainwri	12	6	3	2	1	58.3%	43.4
Duke	14	6	4	2	2	57.1%	42.1
WoodyW	14	6	3	3	2	57.1%	42.1
Capuano	13	4	4	1	4	61.5%	42.1
Suppan	14	6	1	6	1	50.0%	39.8
Zambran	14	6	3	4	1	50.0%	39.8
Belisle	13	6	4	3	0	46.2%	38.9
Bush	13	4	3	4	2	46.2%	37.0
Wells	13	4	4	4	1	38.5%	34.5
Maholm	13	2	3	5	3	38.5%	32.6
Pitcher	STARTS	QS	BS	QIP	QR	WS%	AGW%
Marshall4	4	0	0	0	100.0%	63.4
Jennings5	3	0	1	1	80.0%	51.8
Bailey	1	0	0	0	1	100.0%	50.9
Keisler	3	1	1	0	1	66.7%	44.1
Milton	6	1	2	0	3	66.7%	42.0
Livingsto2	1	1	0	0	50.0%	40.7
Thompso	7	2	2	1	2	57.1%	40.3
Guzman	3	0	1	0	2	66.7%	39.9
Wellemey3	0	1	0	2	66.7%	39.9
Reyes	9	2	2	2	3	55.6%	39.0
Armas	7	1	3	0	3	57.1%	38.5
Chacon	4	1	2	0	1	50.0%	37.5
Albers	7	3	3	1	0	42.9%	37.4
Saarloos2	0	1	0	1	50.0%	34.4
Miller	3	0	2	0	1	33.3%	28.9
Carpente1	0	0	1	0	0.0%	18.0
The last two columns are Winnable Starts Percent (WS%) and Average Game Winnable Percent (AGW%).

WS% is simply the number of QS plus QR starts divided by the total number of starts. IOW the number of starts a pitcher has thrown that results in his team having a greater than 50% chance of winning. (IE Sampson has 11 starts and 10 of them were either QS or QR.) *It should be noted that QS and QR starts have a requirement of 3 or less ER. So you can think of WS% as the percentage of starts in which a starting pitcher has given up 3 or less ER.

AGW% is the percent chance of victory per each game that particular starter threw. The reason that the majority of pitchers are under 50% is due to the fact that the NL-Central, as a whole, has a losing record.

Here are the team numbers:
Code:
TEAM	QS	BS	QIP	QR	WS%	AGW%
HOU	35	11	10	8	67%	46.9
CHC	34	13	9	7	65%	46.1
CIN	32	18	9	6	58%	43.4
MIL	29	12	12	11	63%	44.2
PIT	29	14	10	11	63%	44.2
STL	24	16	11	10	56%	41.2
But

This is just one way of looking at the performance of starting pitchers. It is more results-oriented, and not necessarily skill-oriented. But I still place some faith in archaic concepts such as pitching around errors, pitching out of jams and all that old school nonsense. I think a pitcher can, to some extent, have numbers that might not be that well-received DIPS/FIPS-wise and still be effective.