Selfish, Contributing, Combination and Weak Hitters
In responding to a question posed yesterday by Brock Hanke, I accidentally hit on what may or may not be a useful way of thinking about team offense.
Suppose there is a team, a kind of an average team, which has 1,450 hits, 600 walks/hit batsmen, and 2200 total bases in 5,500 at bats. Such a team would score about 739 runs. Suppose that there is one hitter on that team who hits .263 with a .333 on base percentage and a .400 slugging percentage; we could call that player Troy Tulowitzki, because Tulowitzki in 2008 hit .263 with a .332 on base percentage and a .401 slugging percentage. Let’s give that player 600 at bats.
Suppose, however, that we replace Troy Tulowitzki with a player who hits .315 with a .400 on base percentage and a .400 slugging percentage. That would be Rod Carew in 1982 (.319/.396/.403), so we could say that’s Rod Carew. Or suppose, alternatively, that we take Tulowitzki out of the lineup, but we replace him with a player hitting .263, Tulowitzki’s average, but with more power (.523 slugging) but less on base percentage (.320). That would be Dick Stuart in 1963 (.261/.312/.521) or Nelson Cruz in 2009 (.260/.332/.524). We’ll call this player Nelson Cruz.
Replacing Tulowitzki with either Carew or Nelson Cruz has essentially the same effect. Either substitution would make the team about 22 runs better, moving them up to 761 runs scored. But suppose the team then makes a second similar substitution, replacing a second neutral-type hitter with a second on-base hitter (a second Rod Carew) versus a second power hitter (a second Nelson Cruz.) The second substitution to a Rod Carew type player, because there are more runners now on base, will have MORE effect than the first such substitution. But making a second substitution of a low-average power hitter into the lineup, because there are fewer runners now on base, will have LESS effect than did the first one. The Nelson Cruz type hitter improves the offense by driving in more runs, thus leaving fewer runners on base, but he reduces the number of runners on base in two different ways—first by driving them in, and second by not getting on base himself. The more times you substitute power for runners on base, the less effective the substitution becomes.
To be fully transparent here, as Joe Biden likes to say, this effect is trivial to negligible as long as we are talking about just one or two players. The more you do it, though, the more significant it becomes. There is a law of diminishing returns that applies.
It is clear that we are seeing those diminishing returns. Let us take, for example, all players in history who had 550 to 650 plate appearances, hit 25 to 35 home runs, and had slugging percentages of .450 to .550. In the 1950s there were 60 such players, of whom 29 drove in 100 or more runs. They averaged 608 plate appearances, 23 doubles, 29 homers, 98 RBI. Half of them minus one drove in 100 or more runs.
In the last decade (2010 to 2019) there were 145 such players. They averaged 608 plate appearances, 30 doubles and 29 homers—but only 33 of the 145 players drove in 100 runs That’s 23%. A player now drives in fewer runs than a player with the same number of singles, doubles, triples and homers two generations ago, because there aren’t as many runners on base around him.
Again, we don’t know at this time whether this is a meaningful effect, or merely a documental change. The players in the 1950s drove in 98 runs on average; in the 2010s, they drove in 90. It’s not a huge difference. In the 1950s, only one player in the group drove in less than 75 runs, that being Woody Held, who drove in 71. In the 2010s, 15 of those players drove in less than 75 runs, with the low being 59 RBI. We don’t know whether this means that baseball teams could do better by mixing in more high on-base guys on not; I’m just trying to create a way to discuss the issue.
Which is this. Suppose that we call the low-on-base, high-power hitters "selfish" hitters, meaning they drive in themselves, and the high on-base, low-power hitters "contributing" hitters. That’s what I said yesterday.
Well, OK, but what about players who have BOTH high on base percentages and high slugging? Guys like Mickey Mantle, not that Mantle would be typical of the group. We will call them "combination" hitters, and those who are low in both areas, we will call "weak" hitters. And there will need to be a fifth group, which is guys who are just sort of middle-of-the-pack in both areas. We will call them Centrist Hitters.
If a player has an on base percentage greater than .340 and a slugging percentage less than .425, we will describe him as a "contributing" type of hitter.
If a player has a slugging percentage over .440 and an on base percentage below .330, we describe him as a "selfish" type of hitter.
If a player has an on base percentage greater than .340 and a slugging percentage greater than .450, we will describe him a "combination" hitter.
If a player has an on base percentage under .300 and a slugging percentage under .400, we will describe him as dead space. Just kidding; we will designate him as a weak hitter.
And if a player fits into none of those four groups, we will call him a Centrist type of hitter. We’re actually going to wind up with 6 groups of hitters here.
In baseball history through 2019 there were 13,280 hitters who had 500 or more plate appearances in a season (the file I am working with hasn’t been updated since 2019. Or maybe it was 2020, but nobody from 2020 had 500 plate appearances.)
Of those 13,280 hitters, 2,773 are classified as contributing hitters, or 21%.
1,172 are classified as selfish hitters, or about 9%.
3,892 are "combination" hitters, or 39%.
1,017 are classified as weak hitters or 8%.
3,964 are classified as centrist hitters, or 30%.
462 are classified as not-category hitters, or 3%. Non-category hitters are either hitters who get on base too much to be considered weak hitters, but who have slugging percentages well below the normal range, or players who have very high on base percentages, but not enough power to be classified as combination hitters.
From year to year, players bounce between categories. Craig Biggio had several seasons in which he is classified as a contributing hitter, and some seasons in which he is classified as a selfish hitter. Only "extreme" type hitters are locked into one group.
The most productive player ever in Group 1 (contributing hitters) was Sliding Billy Hamilton, from the 1890s. Other prominent players who had many seasons in this area include Roberto Alomar, Matty Alou, Luke Appling, Richie Ashburn, Craig Biggio, Maxie Bishop, Wade Boggs, Lou Brock, Brett Butler, Eddie Collins, Dom DiMaggio, Ferris Fain, Curt Flood, Nellie Fox, Bret Gardner, Tony Gwynn, Stan Hack, Rickey Henderson, Billy Herman. . .well, you get the idea. But I will also note that Bobby Grich, Carl Yastrzemski, and Darrell Evans also had numerous seasons in this category, since they also had many years in which they were productive hitters but with slugging percentages below .425.
The most productive player ever in Group 2 (selfish hitters) was Alfonso Soriano in 2002 and 2003. Soriano was a leadoff hitter in those years; it is perhaps an odd place in the lineup to place this type of a hitter. Others who represent this group would include Garrett Anderson, Tony Armas, Ernie Banks, George Bell, Andre Dawson, Adrian Beltre, Joe Carter, Willie Davis, Juan Gonzales, Torii Hunter, Dave Kingman, Lance Parrish, sometimes Dave Parker, Salvi Perez and Matt Williams. Frank Thomas, the 1950s guy who just passed away, was almost always in this category.
Almost all of the greatest hitters in history were combination hitters, Group 3, while weak hitters, Group 4, rarely stay in the league long enough to become recognizable names.
But here is what I was trying to get to.
In the 19th century, the ratio of contributing players to selfish players was 339 to 9, or 37 to 1. Baseball was a game in which each player tried to get on base, and each player tried to advance teammates on the bases. It was a team game, in which runs were generated by community actions.
From 1900 to 1919, that ratio was 515 to 15, or 34 to 1.
From 1920 to 1939, the ratio of contributing to selfish hitters was 440 to 28, or 16 to 1.
From 1940 to 1959, the ratio was 390 to 79, or 5 to 1.
From 1960 to 1980, the ratio was 375 to 211, or 1.78 to 1.
From 1980 to 1999, the ratio was 394 to 301, or 1.31 to 1.
From 2000 to 2019, the ratio was 322 to 529, or .61 to 1, or 1.64 to 1 going the other direction. More RBI men than players who are getting on base for them.
I think we have to accept that, unless there is a historic re-direction, a radical change in the trend line, over the next generation the ratio will be 3 to 1 or higher in favor of the selfish type of hitter. And the implications of that are immense. It is not really a "team" game anymore; it is a game of individual actions. This may mean that our offensive formulas don’t work anymore. It may mean that the players that are available and the players that teams need just don’t match anymore; that is, that the Tony Gwynn-style players that would most benefit a typical team just are not anywhere to be found, because players have not been trained to play that way. It may mean that we are miscalculating value in players, because what we figuring in effect is how valuable players WOULD HAVE BEEN to teams 40 years ago, rather than how valuable they are now. It may mean that the game that we are watching is not the game that we grew up watching, the game that we decided to adopt when we were young. And it may mean that the public just doesn’t like the game anymore. They don’t want to watch.
Or not. But it is something that we need to understand better.