I got goosebumps. I literally got goosebumps.

At about 10:30 last night, I had an entire column sketched out. I was watching the Dodgers close out a devastating loss, and after three more outs I was going to write about bad at-bats in the sixth, and questionable resource management in the eighth, and two very good relievers having one very bad night.

Jeff Kent hit a homer. Then, J.D. Drew hit an absolute bomb, and Trevor Hoffman came in. I mostly thought it was nice of Jon Adkins to set up a save situation for Hoffman, and improve the closer’s chance of setting the career record this year. Even the Fox Sports truck was ready for it, posting a graphic about Hoffman’s success against the Dodgers: 55-for-57 in save situations, with no blown saves since 2001.

Then Russell Martin hit Hoffman’s first pitch for a homer. 9-8. Hrm.

Then Marlon Anderson hit Hoffman’s second pitch for a homer, and my skin got all weird.

Hell, it wasn’t even over yet. The Padres, as they had done all night–neither team had a clean 1-2-3 inning in the game–popped the Dodgers back for a run in the top of the tenth. Aaron Sele, who looked like he’d have trouble starting for a good MSBL team, managed to limit the damage to one run. The Pads, who’d also gone through the front lines in the bullpen–hold that thought–turned to Rudy Seanez to protect their lead. That didn’t go well. A Kenny Lofton walk–almost making up for a criminally bad first-pitch, weak-swing double play in the sixth–set up a Nomar Garciaparra two-run homer, a no-doubter into the left-field bleachers, for the most dramatic Dodger win since the 1988 World Series.

Just like that, the goosebumps were back.

It wasn’t just the back-and-forth within this game. The amazing comeback was historic, for one: just the fourth time in history that a team had hit four consecutive home runs. More relevant to the participants was what it did for their fortunes. The Padres had been three outs away from taking a 1 ½-game lead in the NL West race and a 2 ½-game edge on a playoff spot with just 13 games to play. They wouldn’t have clinched anything, but they definitely would have changed the conversation in the NL playoff races. The Dodgers, not the Padres, would have been left fighting for the wild card.

With one…OK, five…swings of the bat, the Dodgers had retaken first place, played the Padres to a draw over the weekend, and re-established a two-game cushion for a playoff spot. What the Dodgers did was amazing, but the importance of the turnaround in the context of the season made it that much more so. The Dodgers now get to play three games against a terrible road team–the Pirates–while the Padres have to go to Phoenix for three against the frisky Diamondbacks, while scoreboard-watching the Phillies, who are playing the Derrek Lee-free Cubs at home.

As I mentioned, I had a number of analytical points I want to make about the game, but most of them seem a bit irrelevant in light of the Dodgers’ comeback. It is worth taking a look at Bruce Bochy’s usage of Hoffman, because even though the veteran closer wasn’t effective on this night, Bochy certainly got as little out of him as possible, a decision that sheds light both on the way in which Hoffman is protected–my argument against his Cy Young candidacy–and the way the scoring rules dictate usage, and sometimes, the outcomes of baseball games.

Hoffman was preparing to pitch the bottom of the ninth during the top of the frame, during which the Padres added three runs to turn a 6-5 lead into 9-5. With no save situation present, Bochy instead turned to Adkins, because Hoffman doesn’t do four-run leads. His 58 opportunities include 44 save situations, and 11 other appearances in tied games. Hoffman has made just three “low-leverage” appearances all season: two getting-work outings in the season’s first two weeks, and protecting a 14-10 lead in the ninth inning at Arizona on May 17.

Think about that: of Hoffman’s last 56 outings, 55 had come with either a save or the game on the line. That’s defining the closer role about as strictly as any manager has ever been able to do.

It might have cost him last night, though, because if there was any time to get out of that box, it was with a chance to step on the Dodgers’ neck. There’s no telling what might have happened had Hoffman started the ninth, of course; remember, his first two pitches were launched into the bleachers. But I have to think that the combination of Adkins and Hoffman was easier to beat than just Hoffman himself.

More frustrating is the decision to take out Hoffman after one inning. It is, again, very typical of Bochy to do so: Hoffman has not recorded more than three outs in a game since August 13 of last season. Still, with Hoffman having thrown just 11 pitches, and the drop-off to waiver bait Rudy Seanez so steep, and the pitchers’ spot due up second in the 11th (before Seanez was inserted into the lineup), this seemed like the spot for an exception. Bochy, instead, stuck with his game plan for Hoffman, and the night was over in a hurry.

It comes back to the argument that we, as an analytical community, have been having for more than a decade. Do defined roles maximize the performance of players in such a way that they make up for not having greater flexibility in usage? While the focus in pitcher usage has been on starters, with so many people concerned that the tracking of pitch counts and the monitoring of workloads has bred a generation of pitchers who can’t pitch deep into games, the real hothouse flowers are relievers. We now have a class of pitchers that expects to throw one inning per game, and in actuality, a sub-class–set-up men–that doesn’t expect to do much more. Relievers, not starters, have been babied to the point of silliness; I can justify the reduced workloads of starting pitchers by looking at changes in the game. I can’t do the same when it comes to relievers.

So Jon Adkins pitches with a four-run lead, and Rudy Seanez pitches in a tied game, and the guy who people want to give awards to and hang plaques for throws 11 pitches in between the two.

And the Dodgers move back into first place.

When last night began, I expected to do my column on the Tigers/White Sox game. The Tigers took care of that for me, though, hitting four home runs and inducing three double plays, cruising to an 8-2 win that just about locks up a postseason berth. At worst, they’ll leave Chicago four games up on the White Sox with 10 games to play, and no more head-to-head matchups. That’s not a clinch, but it’s close enough to start thinking about the playoff rotation.

I had criticized the Kenny Rogers signing in the offseason, but I was wrong about that. Rogers is another in a long line of left-handers who continue to be durable and effective late into their careers. I’ve never though about him as a “Tommy John class” pitcher, but he gets ground balls, kills the running game and generally keeps his home-run rate down. At 41, he’s essentially the same pitcher he was 10 years ago, and he’s been a godsend for the Tigers down the stretch: a 1.80 ERA in 60 innings in August and September, thanks mostly to allowing just three homers in that stretch. In his last two starts against the White Sox, he allowed no runs in 13 innings. He’s not the sole reason the Tigers will go October for the first time in 19 years, but he’s a big part of it.

The difference between the 2005 White Sox and the 2006 version showed up pretty clearly last night. Mark Buehrle gave up three homers in 6 2/3 innings, bringing his total to 34 in 199 1/3 frames. That rate is nearly double his 2005 mark, a jump that far outstrips the small change in his flyball rate. Freddy Garcia and Jon Garland have also allowed more homers in 2006. The drop in effectiveness of the starting rotation can be traced to the higher home-run rates and more balls in play dropping for hits, and that’s been the difference between 99 wins and the 90 or so the Sox will win this year.

That the 2005 team won a title while the 2006 one will probably miss the playoffs will make the difference between the two seem much greater than it actually is. Actually, the two are of roughly equal quality, above-average baseball teams in very competitive divisions. Last year’s team outperformed its indicators by a considerable amount, whereas the 2006 Sox have played just as well as their record indicates. If you want to be pithy about it, you could say that the “luck” they had last year went to the Tigers and Twins in 2006, both of whom are three wins better than indicated. The key point is this: there’s just not that much difference between the two White Sox teams, and the front office shouldn’t allow the emotion of the moment to dictate policy.

One thing to note about the White Sox is that they’re an older team with a middling farm system, certainly the worst or maybe second-worst in the division. Their window for success could close rapidly, making this offseason a critical one for them. They’ve committed to the short term with the Thome trade and the Konerko signing, so they’ll have to continue down that path in an effort to win in 2007. Kenny Williams has been very successful at acquiring major-league contributors, which leads me to believe that the White Sox will be near the top of the AL Central again next season, along with the Indians.