Leadoff man. No. 2 batter. These lineup positions probably bring to mind certain types of hitters -- perhaps a slap-hitting speedster and a guy who "knows how to handle the bat" -- but not necessarily the best hitters.
However, times are changing.
Consider the Cubs' Kristopher Bryant, one of the game's bright young stars and elite batsmen. Bryant has smashed 94 homers and slugged .527 over the past three seasons, yet since 2016, he has made nearly two-thirds of his starts in the No. 2 spot. He recently expressed enthusiasm for the idea of leading off in '18.
Bryant is just one example of a larger trend that fits with how clubs' thinking has evolved through the years, as an analytics-based approach has taken hold.
This trend could be summed up like this: Get the best hitters more opportunities by batting them higher in the lineup.
While the overall effect of lineup structure is relatively small, modern teams are searching for every edge they can find. The 2006 sabermetric bible, "The Book: Playing the Percentages in Baseball" -- co-written by MLB.com's Tom Tango -- established that an optimal batting order would feature a team's three best hitters in some combination of the first, second and fourth spots. The next two best hitters would follow in the third and fifth spots.
To see how these ideas have taken hold, consider the numbers from this group of five seasons, each a decade apart: 1977, '87, '97, 2007 and '17. For each of those years, the best hitter on each team was identified, based on the highest weighted runs created-plus (wRC+), with a minimum of 300 plate appearances for that team. Each hitter's starts for that season were then broken down by batting-order position.
At the broadest level, let's separate the lineup roughly into halves by taking the top four spots and then the bottom five (clubs' best hitters have virtually never hit ninth). In 1977, roughly 72 percent of "best hitter" starts came in the top half of the lineup, but that number has since risen steadily to 87 percent, as seen in the graph below.

For example, in 1977, future Hall of Famers Gary Carter of the Expos (135 wRC+) and Willie McCovey of the Giants (129 wRC+) combined to make about 84 percent of their 279 combined starts batting fifth, sixth or seventh. Meanwhile, the most common Nos. 1 and 2 hitters for both teams each posted a wRC+ below the league average of 100.
In 2017, the only "best hitter" to make 60 percent or more of his starts lower than fourth in the order was Oakland's Yonder Alonso. A breakout performer who discovered his power and launched 28 home runs, Alonso started 70 percent of the time at fifth or lower for the A's before he was traded to the Mariners in early August.
Of course, a club doesn't necessarily know who its best hitters are going to be in advance, and a surprise star such as Alonso can throw a wrench in the Opening Day lineup. The same goes for established stars who underperform. But while it's not completely fair to judge lineup choices in hindsight, it's evident that teams are doing a better job of first identifying their best hitters -- wRC+ and other advanced stats didn't exist until recently, after all -- and then deploying those hitters in an efficient manner.
Looking deeper than the top and bottom halves of the lineup, the percentage of "best hitter" starts coming in the top two spots has jumped from 11.3 percent in 1977 to 17.2 percent in 2007 to 28.2 percent last year. Conversely, the percentage of those starts coming in the fifth and sixth spots has tumbled from 25.3 percent to 11.5 percent over that same period.
The results show, in particular, a recent explosion in "best hitters" occupying the No. 2 position. In 1977, '87, '97 and 2007, these players got somewhere between 2.7 percent and 6.4 percent of their starts batting second. This past season, that number was 15.5 percent.
Leading the charge as No. 2 hitters were Bryant and the Marlins' Giancarlo Stanton (110 starts apiece), with Stanton, the National League Most Valuable Player Award winner, setting a single-season record with 47 home runs out of the No. 2 spot before getting traded to the Yankees in the offseason. The Cardinals' Tommy Pham (87 starts), the Blue Jays' Josh Donaldson (74) and the Angels' Michael Trout (71) also saw significant time batting second.

Another way to view the same phenomenon is by examining how the overall performance of different lineup positions has changed over time.
In 2017, all hitters in the top two spots in the order combined for a .771 OPS, compared to the overall MLB mark of .750. That OPS gap of plus-21 points was the largest for any season going back to 1977 -- tied with '16 and '15. Meanwhile, '14 trailed not far behind (plus-15).
As the chart below demonstrates, Nos. 1 and 2 hitters have generally fared little better than -- and sometimes significantly worse than -- the average hitter. That dynamic has changed dramatically in the past four seasons.

Of course, exceptions remain. While the Rockies' Charlie Blackmon, the Twins' James Dozier and the Astros' George Springer each hit at least 34 homers in the leadoff spot in 2017, the Reds chose Billy Hamilton for the same role. The lightning-fast center fielder made each of his 135 starts in the top spot, with Cincinnati focusing more on his baserunning (59 steals) than his hitting (.247/.299/.335).
Still, the future might look more like a late-June series from last season between the Marlins and Cubs. In two of the games, Stanton batted second for Miami, while Chicago opened its lineup with the one-two punch of Anthony Rizzo and Bryant. Those three sluggers combined for 120 homers, and each ranked among MLB's top 30 qualified hitters in wRC+.
With lineups like those, terms such as "leadoff man" and "No. 2 batter" could be coming to mean something significantly different than they have for much of baseball history.