On Saturday, Mets center fielder Carlos Beltran hit two home runs against the Arizona Diamondbacks. The first was a first- inning line drive that put the Mets up 3-0, and the second was a seventh-inning line drive that put them up 7-5, helping quell a furious Snakes comeback. The second bomb was Beltran’s 17th home run of the season, and with it he surpassed last season’s total in about a third as many at-bats.

Beltran, who’s playing the field and running the bases with his usual flair and intelligence, is in the middle of the best season of his career and on pace to destroy the Mets’ single-season record for home runs, which belongs to Todd Hundley and his 41 homers in 1996. As the single biggest reason the Mets are in first place, Beltran is also, with the injury to the Cardinals’ Albert Pujols, in the race for the National League MVP award.

This is a pretty startling turn. Last season was the worst of Beltran’s career, as he batted .266 BA/.330 OBA/.414 SLG. Most were willing to give him a pass – he played through a series of nasty injuries – but more troubling than the results was his passive approach. Bunting for base hits, taking pitches in hitter’s counts and swinging in pitcher’s counts, and looking tentative and confused, he didn’t look like anyone’s idea of a franchise player. When he opened this season 0-for-9, he really started to hear it from the fans, and when they demanded a curtain call after his first hit (a long home run), Julio Franco had to talk a visibly frustrated Beltran into giving it.

What’s happened since to bring Beltran back to form as an elite power hitter? There are as many theories as you’d care to hear, the most popular being that the imposing presence of Carlos Delgado both in the lineup and in the clubhouse has allowed the reticent Beltran to play without a sense of suffocating pressure. You can’t discount that, but ultimately the answer is unknowable, even to Beltran: Sometimes you’re going to hit, sometimes you aren’t. His performance is so fundamentally different from last year’s that the more interesting question is how – rather than why – he’s improved so much.

The first obvious difference is that Beltran’s been hitting in better counts, and working longer at-bats. Last year, he walked 56 times in 582 at-bats; this year, he’s walked 36 times in 195 at-bats. The walks are, of course, good in their own right, especially with Delgado and David Wright lurking behind him in the lineup, but they’re a byproduct of a more patient approach, one that’s visibly different from last year’s. Instead of trying to make things happen on every pitch, Beltran is waiting for good fastballs and smoking them. He actually isn’t hitting breaking pitches or change-ups particularly well, according to my scorecards, but if he avoids offering at them, he doesn’t have to.

The second, less obvious difference is that Beltran is hitting the ball in the air more and it’s going over the fence a lot more when he does. Last year, according to data at hardballtimes.com, 44% of the balls he put in play were grounders; that number is 37% this year. The really shocking number, though, is that last year 9% of the fly balls he hit cleared the fences; this year, 25% of them are leaving the yard. That number is ridiculous, and almost certainly unsustainable, but along with the walk totals it paints a pretty accurate description of what Beltran’s been doing well: focusing on the high fastballs he can lay a short, fast swing on, and driving them with immense power when he sees them.

Add in that ability to bunt for base hits (which is quite impressive if it comes in the context of MVP-level production) and his penchant for timely, dramatic hits, and you have the complete offensive player everyone expected when Beltran signed after the 2004 season.

The big question is whether he can keep this up. That’s easy to answer – if he can play like he did last year for a full season, he can certainly play like this for a full season. This season’s performance is, if anything, closer to his true level of ability. Last year, there was something like a 10% chance that he would play as badly as he did, and a 20% chance that he would play as well as he’s playing now; the same was true coming into this season. Performance is highly variable – players change their approaches and batting stances, see the ball well or poorly, hit in good or bad luck, and all we can do is take it as an indicator of how good they are. There’s nothing, in other words, to read into the fact that Beltran did poorly last year and is doing spectacularly well this year; nothing’s changed, and he’s still the same player.

This can be one of the more frustrating things about baseball in some ways, but its unpredictability is also an immense part of its charm. Don’t expect Beltran to hit 50 home runs this year, but don’t be surprised if he does.

tmarchman@nysun.com