On a good night, Zach Duke can out-duel the likes of Cy Young-winners Johann Santana or Roy Halladay. On a bad night, well, last night’s nine-run outing against the Florida Marlins was all too “good” an example of such.

Do these extremes net out, or is one more important than the other in determining Duke’s worth to the Pirates?

Matt Shetler nailed it when he suggested that Duke be NON-tendered and not brought back in 2011. While Duke isn’t exactly the worst pitcher around, Shetler rightly pointed out that Duke IS about as bad as a VETERAN pitcher will likely be.

Put another way, someone with Duke’s experience should be decidedly better than struggling rookies and near-rookies like Brad Lincoln, Dan McCutchen, and Charlie Morton. Except that he isn’t.

Age might not be the issue here; length of service is. Brian Burres is actually older than Duke. But in addition to making much less than Duke, Burres appears to be a late bloomer with less experience who has finally hit his stride and is on an upcurve appropriate for someone of his experience level (Although he’d be more convincing if it had taken place five years earlier).

On the other hand, Duke was an early bloomer that entered a long term decline. Duke was the Pirates’ great hope as of 2005, when he got off to a sensational start (an ERA of 1.81 for a half season).

But even then there were signs that things weren’t as good as they seemed. At that time, Duke’s  FIP (a “sabermetric” ERA estimated by home runs, walks, and strikeouts), was more like 3.00. That suggested that he was a strong pitcher, but not the Cy Young contender he initially appeared to be.

Then 2006, 2007, and 2010 were all worse than the preceding years. This pattern was interrupted by an aborted rebound in late 2008 and early 2009. Injuries over the year took their toll.

But the main problem is that Duke has less speed and power than the average pitcher, so he needs to use finesse and “location” to get outs. This was an advantage in the early going, when no one knew him. But he’s now a tired subject, because unless he’s at his very best, he’s eminently hittable.

So where does Duke now stand? Based on ERAs to date, the Pirates’ 2011 rotation should be James McDonald, Ross Ohlendorf, Jeff Karstens, Brian Burres, and Paul Maholm. Maholm makes the cut, Duke doesn’t (If one used FIPs, one would put Maholm ahead of Karstens and Burres, and possibly Ohlendorf).

As a third year “arbitration” player, someone of Duke’s seniority would probably command a salary of over $5 million. That would suggest that he’s a serviceable, though moderately below average starter. But he’s actually now of “replacement,” not below average, caliber. A low budget team like the Pirates can’t afford this from one of their highest paid players.

Suppose Duke were willing to take a pay cut to $1-2 million. Could he be brought back as a reliever/situational starter? Not really, because his profile is all wrong.

Some of the weaker Pirate pitchers such as Dan McCutchen, Jeff Karstens, and even (to a certain extent) Paul Maholm, have what I call a “last inning problem.” That is, they will pitch well for X innings, then crumble in the X plus first. In that case, managing them consists of determining what X is, and keeping them within that limit.

Duke, on the other hand, has a “first” inning problem, the first could well be his worst. If he survives it and settles down, he can pitch well for most of the remainder of the game. Last night, he didn’t.

In a starter, this is acceptable in a mild form (the Duke of early 2009). He’ll give up one or two runs in the first two innings, no more than one more by the sixth, and possibly work the seventh or eight.

But this is unacceptable in a reliever, for whom the first inning is usually the last inning. If such a person gives up a run, say every other time he takes the mound, that leads to probable defeat.

Duke was the star of a promising but ill-fated (2006) rotation consisting of himself, Paul Maholm, Ian Snell and Tom Gorzelanny. The latter two had to be sent down to the minors, then traded away (although Gorzelanny is now prospering in his home town of Chicago).

It’s now time to say goodbye to Duke (and absorb Maholm into the new rotation mentioned above.) In doing so, the Pirates will say goodbye to what has been a major liability for the past half decade, and hopefully move on to a new start.

Read more MLB news on BleacherReport.com