Roger Clemens is back, treading the same path that Elvis Presley, Frank Sinatra, and George Lucas walked before. Clemens aborted his retirement after just a few months, returning to the Houston Astros yesterday in exchange for a minor league contract that pays $322,000 over the five-month minor league season, followed by a prorated percentage of a one-year major league contract worth $22,000,022 when (and, since all things are possible including a Clemens flame out, if) he is added to the big league roster.

As Presley, Sinatra, and Lucas proved, comebacks and sequels are marked at the beginning by great expectations, but often prove to be only inadequate reprises of the original. The audience descends into disappointment and then apathy. Perhaps these artists even produced some good work during their second time around, but competing with your own legend when you’ve been at the pinnacle of your profession is a quixotic task at best.

This is especially true for Clemens, who at age 42 went where no pitcher had gone before.In the entire history of baseball,few pitchers have had seasons as dominant in their 20s, let alone their 40s, as Clemens did in 2005. In the period spanning the last 105 years, no pitchers aged 40 or older have come close to Clemens’s 1.87 ERA against a league average of 4.23. Randy Johnson, who posted a 2.60 ERA against a league mark of 4.31 as a 40-year-old in 2004, came the closest, but the Big Unit’s marks weren’t at the same immortal level as the Rocket’s.

The gap between Clemens’s ERA and the league’s was the seventh-best in history (175 innings or more).In the seasons in which the six wider differentials were posted (by just four pitchings – Pedro Martinez, Greg Maddux, Dazzy Vance, and Clemens himself in 1997) all of the pitchers were younger than Clemens – Martinez did it at 27 and 28, Maddux at 28 and 29. Clemens was 34. Only Dazzy Vance, at 39, could be said to have been “old” at the time.

Normally, when an athlete misses significant time at an advanced age, he finds his body has difficulty responding when he gets back into the harness. Reflexes decay from inactivity, muscles go slack and won’t tense up again. This seems to be less of a worry for Clemens, who pitched in the World Baseball Clas sic and by all accounts has been keeping himself in good working order during his layoff.

There are two unknowns, then, in Clemens’s return. The first is age. Most 40-somethings don’t pitch like Clemens does. In fact, they don’t pitch. Just because Clemens has defied the aging curve to this point doesn’t mean it won’t suddenly catch up with him.Take Randy Johnson as your example here.

More seriously, it’s not just Clemens’s age that makes his return an unlikely high-wire act, but his very excellence. Earned run averages under 2.00 just aren’t normal, not even for Clemens, whose career ERA is 3.12.The term “unlikely”really doesn’t do the possibility of a repeat justice. If Clemens’s 2005 was a minor miracle, a repeat would be a major one.

A third possible complication, one not unique to Clemens but common to all older pitchers, is the possibility that he might suffer the sudden, crippling arm injury that he has avoided for so long. That’s how Nolan Ryan’s string was finally cut in 1993, when he was 46. One moment he was an uncommonly durable pitcher, the next he was a banker being courted by Texas Republicans.

As for the Astros, this could be a season-changing acquisition even if Clemens is not great but just competent. At this writing, the Astros are a .500 team, which is actually a bit better than they should be given their runs scored and allowed. The offense is mediocre. Morgan Ensberg has hit well, as has Lance Berkman, though the latter has often missed time with injuries. Craig Biggio has outperformed his age, Mike Lamb has rebounded from a horrific year to excel in a part-time role, and Brad Ausmus, at .298 AVG/.386 OBA/.358 SLG, is rapidly coming back to earth after an unusually hot start.

Those are the high points.Ex-Met Preston Wilson has been a massive disappointment, with just five home runs in nearly 200 at bats, and Jason Lane (.194/.326/.328) has been a disaster only slightly mitigated by his nine home runs and 32 walks. The offense doesn’t have enough oomph to bail out the team when the pitching doesn’t perform.

And for the most part, it hasn’t performed. Traditionally a strong suit of the Astros, the bright spots have largely been limited to Roy Oswalt and Wandy Rodriguez, the latter of whom is also returning to his 5.00-plus ERA ways after a strong April. Andy Pettitte has regressed as the season has progressed, and closer Brad Lidge has been mysteriously ineffective.

As presently constituted, the Astros won’t be good enough to make up the 6.5 games that currently separate them from the NL Central-leading Cardinals, or even win the wild card. Not only will Houston most likely fail to defend its NL pennant, but there are likely rougher times ahead. A team’s expected won-lost record, extrapolated from runs scored and allowed, is a more accurate predictor of a team’s future than it’s literal won-lost record. The Astros have scored 245 runs and allowed 266. This would normally net a team a losing record, and if the Astros continue in this vein, it ultimately will.

Even the spectacular Roger Clemens of last season probably won’t be enough to push the Astros past their limitations, so it’s possible they will spend a great deal of money on something that will have little more meaning to the franchise than one more glorified farewell tour.

Mr. Goldman writes the Pinstriped Bible for www.yesnetwork.com and is the author of “Forging Genius,” a biography of Casey Stengel.