It struck me last week, not long after Jose Castillo sent another ball over a fence, sending Freddy Sanchez home ahead of him, that the Pirates should win more than they lose the rest of the way.

Not a winning season, just more wins than losses after Memorial Day. That modest prediction is not based on the recent 7-3 homestand. It comes from improved starting pitching and the far right column of the box scores.

I'm not used to seeing as many as six Pirates in the lineup above .300. A Pirates fan from the Bronx, Corey Bunje Bower, noticed that, too, and calculated the combined season batting average of the eight starters who put up 26 runs in three nights against the Brewers. It was .306 after more than 1,100 at-bats.

Plate appearances that had gone to Joe Randa, Chris Duffy and Humberto Cota in April have been assumed by Freddy Sanchez, Jose Bautista, Nate McLouth and Ronny Paulino. Thus chronic dead spots have disappeared. Check the team improvement in batting average, on-base average and slugging from April to May, resulting in almost two more runs per game.

That's not all because of lineup changes. Jason Bay, probably baseball's best outfielder since the start of 2005, was National League player of the month (12 HR, 35 RBI, .321/.402/.688). Castillo went from one of the league's weakest hitting second basemen to one of the strongest with a torrid .366/.413/.634 in May.

We'll see if Castillo and Sanchez stay among the top hitters at their positions, but Sanchez now is hitting .302/.343/.420 in 646 career AB as a Pirate. Any talk of moving him from third should be met with swordplay.

Less is known about Bautista, McLouth and Paulino, but the improvement at their positions is undeniable. Just scan the numbers generated by Pirates while playing catcher, third base and center field this season, particularly their on-base and slugging averages.

McLouth's numbers include a miserable April when he faced a lot of left-handed pitching. He had a good May (.263/.362/.413 with 20 runs scored) and has seven steals in eight attempts.

Right field should improve, too, if Jeromy Burnitz never hits against a left-handed pitcher again. Few hitters this season have hit lefties as well as Craig Wilson (.349/.451/.674 in 43 AB) or as badly as Burnitz (.184/.231/.286 in 49 AB.)

Given that Wilson and Burnitz are roughly equivalent against right-handed pitchers, and one or both could be traded. If either one is traded, more playing time should go to Bautista and McLouth.

Anyway, the team was up to .263/.325/.421 through Sunday, nearly abreast with the league .263/.334/.424. The relatively weak OBA stems from being next to last in taking walks.

I haven't much space left for pitching, but the improvement is also easy to see. Opposing teams hit a ridiculous .297/.375/.485 with 36 home runs in 26 games in April, but were held to a manageable .268/.347/.392 in May. Pirates pitchers gave up only 22 home runs in the 27 games.

The strikeout-to-walk ratio went from 1.49 in April to an above-average 1.98 in May. Hence the drop in earned run average from 5.30 in April to 3.89 in May. Pirates pitchers' season ERA of 4.53 through Sunday is now nearly as low as the league mark of 4.39.

That is mostly due to improvement by the young starters. A Pirates fan in Murrysville, Ron McClure, has traced their progress through quality starts.

A pitcher gets a quality start if he pitches six innings and gives up three earned runs or fewer. In April, only 10 of the 26 starts were quality. Since May 1, 16 of the 31 starts have been.

No one could call it a great rotation. It's more "Jeff Suppan by Committee," as one friend put it, but it's moving in a positive direction, and Tom Gorzelanny is waiting with a 2.72 ERA in Class AAA Indianapolis.

It seems only the Pirates' freakish inability to win one-run games left them 12-15 in May, despite scoring 27 more runs than their opponents in the 27-game month.

So what does it all mean? If the Pirates finish with four decent to good starting pitchers, above average players at second, third, short and in left field, an average catcher and a workable solution in center field, a smart organization should be able to buy one or two impact players with the $20 million or more dropping from the payroll.

Then we could talk about winning seasons.