Recent success of a few scattered Pirates brings to mind an old joke that goes like this:

First Guy: "I'm an out-of-work comedian. Ask me what my problem is."

Second Guy: "What's your ..."

First Guy: "Timing!"

OK, it's not a great joke, but this is not a great team. We work with what we have. So let's look at Tony Armas' transformation from pitching disaster to steady thrower at a time when it's bound to do the Pirates the least good.

I was among the few who liked Armas' signing for $3 million last winter. (Tack on a $500,000 buyout if the team doesn't pick up Armas' $5 million option in 2008, which it assuredly will not.) Armas' task here was simplicity itself: Don't be awful. He was succeeding a flailing foursome (Victor Santos, Shawn Chacon, Oliver Perez and Kip Wells) who had combined for a 5.87 earned run average in 50 starts in 2006.

A 29-year-old pitcher with a 4.45 career ERA, and a 5.03 ERA in 2006, Armas seemed a good bet to stay in that neighborhood, which is less crowded than you might think. According to The Hardball Times, the average ERA for No. 4 starters in the NL last year was 5.11 and the average No. 5 ERA was 6.26. There isn't enough talent to fill out most five-man rotations.

Armas was beyond awful in April. By the time he was banished to the bullpen in mid-May, he was 0-3 with an 8.46 ERA, though the Pirates had somehow won two of his seven starts.

Early in July, Armas figured something out. His ERA last month was 0.75 in 12 innings, with 12 strikeouts, 2 walks and 5 hits. He got a start Aug. 1, the day after the Pirates traded for Matt Morris, a past-his-prime pitcher who will make $9.5 million next season in the hope he can be a stable, league-average, innings-eater.

That's like the Armas bet again, but to a higher power. Morris, 32, is a right-handed, low-strikeout pitcher, not a great fit in PNC Park. If he can take the ball every fifth day and maintain an ERA around his current 4.53, he'll be OK. But the trend is not his friend: Morris has not won since June 11, his July ERA was 8.48, and it's 6.57 after two starts with the Pirates.

Armas, of course, is 2-0 with a 2.92 ERA in August, with 7 strikeouts, 2 walks and 10 hits (two home runs) in 12 1/3 innings. The Pirates are 4-5 in his starts this season, a .444 percentage just a hair better than the team's overall record.

The Pirates can't keep Armas, even if he finishes with a decent record. They have Tom Gor-zelanny, Ian Snell, Paul Maholm and now Morris. They also have Zach Duke and Shane Youman. They have Bryan Bullington, John Van Benschoten and Sean Burnett. That is, if not "an embarrassment of OKness" as one friend put it, a slew of young pitchers who need their shots at the rotation.

Armas might help the Pirates win a few more games but is no building block. Maybe he'll clear waivers and go to a contender that needs him for a spot start or to fill out the bullpen, but the Pirates would not get much for him and it would be a bitter irony if Armas outpitches Morris.

The other piece of bad timing is outfielder Xavier Nady's injury. Converted catcher Ryan Doumit has taken most of the starts in right field since and is doing OK there. But it's not Nady's bat that needs replacing. It's catcher Ron Paulino's. Through Sunday, Paulino had hit .190 with a .239 on-base average and .310 slugging average against right-handed pitchers, while slamming .374/.418/.527 in 98 plate appearances against left-handers.

Doumit, a switch-hitter, seems the ideal platoon complement, hitting right-handers at .289/.355/.532 but only .255/.333/.333 against left-handers.

Trouble is scouts have a big problem with Doumit's catching, and the Pirates' record when he catches (6-22 through Sunday vs. 41-40 for Paulino) does nothing to show they're wrong. Paulino is said to call a better game, but a check of pitching splits on baseball-reference.com doesn't turn up much evidence.

Pirates opponents are hitting .280/.341/.430 when Paulino catches, and .281/.351/.453 when Doumit catches. That's close. Opponents have been more likely to bop a home run or draw a walk when Doumit has caught, but his catching splits have been improving these past two months as Paulino's have slid. That doesn't mean the scouts are wrong, but when the splits don't show a night-and-day difference and Doumit's value as a catcher is higher, the end of a lost season seems a good time to find out once and for all if Doumit can hack it behind the plate. Finding out he can play right field without embarrassing anyone is not nearly as helpful.

What's the use of screwing up another season if the Pirates can't find out more about the players they're keeping?

Timing -- did I mention it has been off since October 1992?