Starting pitchers are like your friends with wives (or serious girlfriends).  When you invite them over for poker, they always have to check with their wife first.  You know you have that friend that everyone likes to hang out with but his wife always keeps him from hanging out with the guys.  It's tempting to plan a night of poker with 5 friends.  What a great time you'll all have drinking beer, farting without fear and making fun of each other.  But what happens when Mike calls and says his wife has planned a night at the theatre?

Well, poker with 4 people just isn't the same.  You need five to have a good game.  Someone is bound to go out early and then you have the guy who only plays the good hands and you are left with the same two guys betting against each other every hand.

So, back to Mike.  Well, now you only have 4 guys to play poker and you don't want to do that.  So, you have two options: you can call your younger brother or you can call your boss who has been asking about getting in on the game.  Either way you are taking a chance.  Maybe your younger brother just isn't mature enough to hang out with the guys, gets too drunk and makes a fool out of himself.  You'd like to bring him into your group, but it may be too soon.  He needs to get that college life behind him first.  Your boss has been cool ever since you met him and you're pretty sure he will fit in well.  But he will pose some problems.  Your little brother might get pissed that you didn't invite him.  You have to watch what you say around your boss as well, or you could create major problems at work.  And as much as you think your boss would fit in, he only drinks bottled beer from Belgium.

Now, what if you had six friends in your crew and you invited them all?  Six is one too many to play poker, though.  It takes too long, someone always takes forever to bet and that makes the other guys mad.  But what are the chances that everyone would come every week?  The chances are better that most nights 5 guys will come.  Some nights all six will come and some only four will come.  But most of the time, five guys will be there.  What makes me say that all six won't come?  Because you can never count on guys with wives!  To bring it back to baseball, here are the Phillies' starting pitching stats for the last 5 years, including our start of the season rotation (original 5 friends) and trades (boss).  The rest of the starting pitchers we used were callups or bullpen arms (you and your friends' little brothers*).

2006 GS & ERA (as starter)
Myers      31    3.91
Lieber      27    4.93
Hamels    23    4.08
Lidle        21    4.74
Madson   17    6.28
Wolf        12    5.56
Floyd       11    7.29*
Mathieson 8    8.27*
Moyer       8    4.03
Brito         2    10.38*
Bernero    1     36.00*
Fultz         1    16.20*

Team Starter's ERA 5.08 (14)
4th most innings pitched (921), most runs allowed in NL (569)

We started 2006 with a rotation of Myers, Lieber, Lidle, Hamels & Madson
We traded for Moyer and traded away Lidle.

2005 GS & ERA (as starter)
Lieber   35   4.30
Myers   34   3.72
Lidle     31   4.53
Padilla  27   4.71
Tejeda  13   2.87*
Wolf     13   4.39
Brito     5     3.80*
Floyd    4     6.75

2005 team starter's ERA 4.20 (9)
10 in IP (957)

We started 2005 with a rotation of Lieber, Myers, Wolf, Lidle & Floyd. Padilla on DL.
We made no trades for starting pitchers.

2004 GS & ERA (as starter)

Milton      34    4.75
Myers       31    5.50
Millwood   25    4.85
Wolf         23    4.28
Padilla      20    4.53
Lidle        10     3.90
Abbott      10    6.24
Floyd        4     3.80*
Powell       2    4.15*
Hancock    2   11.57*
Madson     1    81.00*

2004 team starter's ERA 4.91 (13)
10 in IP (922)

We started 2004 with a rotation of Millwood, Wolf, Milton, Padilla & Myers.
We traded for Lidle and picked up Abbott as an in-season FA.

2003 GS & ERA (as starter)
Millwood     35    4.01
Padilla        32   3.62
Wolf           33    4.23
Myers         32    4.43
Duckworth  18   5.04
Telemaco     8     3.97*
Roa              3    7.50*
Silva            1     6.75*

2003 team starter's ERA 4.20 (8)
8 in IP(969)

We started 2003 with a rotation of Millwood, Padilla, Wolf, Myers & Duckworth.
We didn't trade for a starter.

2002 GS & team starter's ERA
Padilla          32    3.28
Wolf             31    3.20
Duckworth    29    5.42
Adams          19    5.00
Person          16    5.44
Myers           12    4.25*
Roa               11    4.22*
Coggin          7     8.13*
Mercado       3     3.55*
Junge            1     0.00*
Timlin           1     couldn't find*

2002 team starter's ERA  4.28 (10)
8 in IP(949)

We started 2002 with a rotation of Person, Wolf, Padilla, Duckworth & Adams.
We didn't trade for any starters in season.

What we can see from these stats is a correlation between the number of guys brought up from the minors and the starting pitchers cumulative ERA.  The more guys we bring up, the higher the ERA for the starters.  2002 would be the exception to this rule.  Both Myers and Joe Roa came up and pitched very well.  It is very hard to predict, however, who will come up from the minors and pitch well.

In 2003, a lot was expected from Duckworth and not as much from Telemaco.  In 2004, Floyd and Madson had some expectations while Hancock and Powell didn't.  Same for Floyd in 2005 with less expected from Tejeda.  In 2006, big prospects Floyd & Matheison were brought up while Brito and Bernero were brought up with little fanfare.  You just can't tell who will do well.

Each year, there is a pitcher who is brought up for a few (3-5) spot starts:
2002 with Mercado, 3 @ 3.55
2003 with Roa, 3 @ 7.50
2004 with Floyd, 4 @ 3.80
2005 with Brito, 5 @ 3.80
2006 with Brito/Bernero, 3 @ 10+

Candidates for this role in 2007 (in order of my preference) would be Germano, Floyd, Happ, Segovia and Brito.

We have obtained 3 starting pitchers via in-season trade or release in the last 5 years (Lidle, Abbott, Moyer).  Lidle and Moyer worked out well while Abbott busted.  This gives us some reason to believe that the team can find a quality starter before the trading deadline if necessary.

Back to what we know based on 5 years data...  Bringing up less guys from the minors equals a better ERA for the starting pitchers, both in real numbers and in comparison to the rest of the league.  We can also look at Games Started the last five years for additional guidance.

Games started by our opening day rotation by year:
2002 (32, 31, 29, 19, 16) 4.28 (10th)
2003 (35, 33, 33, 32, 18) 4.20 (8th)
2004 (34, 31, 25, 21, 20) 4.91 (13th)
2005 (35, 34, 31, 27, 13) 4.20 (9th)*
2006 (31, 27, 23, 21, 17) 5.08 (14th)

*  I used Padilla as a starter instead of Floyd, since Padilla was on the DL at the start of the year, but expected to be part of the regular rotation.

In the best two years of 2003 and 2005, we got 33 or more and 31 or more starts, respectively, from our top three pitchers.  In those years also, the starter with the fourth most starts had more than the starter with the third most starts in 2004 and 2006.  The number of starts for our number 5 guy is not an accurate predictor of the teams starting ERA.  When we have 4 of our top 5 guys starting a lot of games, we do well.

I see the strongest correlation between the 5 starters we have now and the 5 starters we had in 2004.  Both years we had 4 guys coming back from good seasons as well as a FA back-end starter (although Milton was more of an innings-eater than Eaton).  Even in a year where we were more confident in the 5 starters we had (2004), we still had to trade for 2 starters.  My question would be if you knew what was going to happen in 2004, and you had Lidle to start the year, would you have traded him before the season started or kept him as a hedge against injuries?

I submit that in order to have 4 quality starters who pitch 27 games or more, we need 6 quality starters to begin the season.