The Perils of the Three-True-Outcome Slugger

For patient sluggers, strikeouts are a necessary byproduct of an approach designed to draw walks and hit home runs. We are well past the point of ridiculing strikeouts as bad when there are tangible trade-offs in on-base percentage and slugging. Those trade-offs are generally good for the player, but for players like Chris Carter this season, the three-true-outcome approach can go very wrong if the power drops or if the (relatively few) balls in play are not falling. For Chris Carter, a drop in both power and BABIP has resulted in a below replacement-level season despite leading the the league in Three True Outcomes.

Trying to get walks and home runs is generally a good strategy for hitters. Adding strikeouts to the mix is fine as evidenced by the leaders in Three True Outcome percentage (HR+K+BB/PA) below:

	Team	PA	HR	BB	SO	TTO%	wRC+	WAR
Chris Carter	Astros	388	17	51	128	50.5	87	-0.3
Joc Pederson	Dodgers	476	23	77	140	50.4	127	2.8
Chris Davis	Orioles	481	34	53	151	49.5	141	3.6
Kris Bryant	Cubs	467	16	62	144	47.5	124	3.9
Bryce Harper	Nationals	473	30	88	97	45.5	196	6.8
Paul Goldschmidt	Diamondbacks	520	22	92	106	42.3	169	5.8
Mike Trout	Angels	499	33	61	116	42.1	171	6.7
J.D. Martinez	Tigers	482	31	38	133	41.9	144	4.7
Lucas Duda	Mets	463	21	50	122	41.7	130	2.5
Adam LaRoche	White Sox	415	10	43	119	41.4	77	-1.1
Mike Napoli	– – –	391	13	46	103	41.4	86	0.1
Joey Votto	Reds	507	21	93	93	40.8	167	5.2
Brandon Belt	Giants	447	17	44	120	40.5	139	3.6

For Carter, leading the list this season has not been a boon to his overall statistics. Carter has a .225 BABIP that is 50 points below his career average, resulting in a .180 batting average — and even a robust 13% walk rate could only get him to a .296 on-base percentage. While 17 home runs is a decent numbers, Carter averaged 33 home runs per 600 plate appearances in the two-and-a-half seasons entering this one, including 37 home runs last year. Houston’s home ballpark should be a help to a right-handed power hitter like Carter, but his pull percentage, hovering around 50% for his career, has dropped all the way down to 39% this season. He has been vulnerable against the fastball and is swinging through more pitches in the zone, unable to take advantage of the close left field fence in Minute Maid Park.

Carter is not alone in his struggles. Adam LaRoche has has had a rough year. Brandon Moss, barely off the list above, is struggling. Joc Pederson got off to an incredible start, but has been only experiencing the worst of the three true outcomes for a short period after the All-Star break. Given Carter’s recent success and the generally excellent numbers put up by many other TTO players, I was curious about whether these players were more prone to having bad seasons or getting worse sooner than the players who are successful without such a large percentage of their plate appearance ending in a strikeout, walk or home run.

I ran a small experiment, looking at qualified hitters from 2008 to -10. I isolated above average (100 wRC+ or higher) players whose TTO percentage was greater than 40% and who also received playing time from 2011 to -13. I ended up with eight players who had an average wRC+ of 124 and 8.7 WAR over the three years. I also took a group right in the middle in terms of TTO percentage using the same criteria above. I was able to find a larger group, of 15 players, with a wRC+ of 120 and an average of 8.3 WAR over the three years. I did not control for age, but the TTO group’s average age in 2011 was a bit older (29.6) due to a higher portion of players in their early 30s compared to the non-TTO group’s average age (28.8).

Looking at the period from 2011 to -13, the TTO group maintained their offense reasonably well, producing a wRC+ of 115, although the average WAR over that period dropped to five wins above replacement. The non-TTO group slightly improved their wRC+, moving up to 123 and maintaining their average WAR at 8.6 over the three-year period. The three true outcomes are often associated with old-man skills, so the decline in WAR could have been the result of a steeper decline, but the results of this small exercise are hardly definitive.

The last exercise made me curious about how closely tied a player’s offensive value is to his overall WAR. Looking at the players on the leaderboard above, ten of the top thirteen players above have a wRC+ above 120 and those players all had solid WAR totals. The graph below shows all qualified batters’ WAR total on the y-axis and the wRC+ total on x-axis. Players with more than 40% TTO percentages are in orange.

For the players right around the line moving through the graph, their offensive production is right in line with their total value. For those above it, they are likely providing a solid defensive value to supplement their offense, and those below the line are not likely much help on the defensive side. In general, WAR and wRC+ correlate fairly well, but for the TTO population (18 batters), the r-squared was a very high .93, necessitating further study below.

Also of note: there were no players in this season’s TTO group with a wRC+ between 90 and 119, so I decided to broaden the sample a bit to include more players to see whether this season is an anomaly. Looking at all players from 2010 to -15, I found that there were plenty of players with a high TTO percentage who fell between 90 and 119, although not quite as high of a percentage as MLB players on the whole. The chart below shows the percentage of players in each wRC+ group, both for TTO hitters and MLB players as a whole.

wRC+	Below 90	90-119	120+
TTO	10.0%	35.7%	54.2%
MLB	17.9%	47.5%	34.6%

Generally, full-time TTO players have performed better than the rest of MLB, recording an average per player wRC+ of 122 compared to 111 for all full-time players. The disparity in WAR per player does not correspond with the difference in offense, however: while TTO players have produced a 2.6 WAR on average, full-timers have averaged a 2.9 mark. The difference in offense without the corresponding difference in WAR was curious.

The same graph from above, but repeated for all players from 2010 to -14, looks like this:

The graph appears to show a lot more orange underneath the line rather than above it. For the population of more than 700 qualified players from 2010 to -14, the r-squared between wRC+ and WAR was .54, not quite as strong as it had been for 2015 alone, but still fairly well correlated. There were 53 players in the TTO group during this period, and the graph for only those players’ WAR and wRC+ is below.

This might not come as much of a surprise, but the value of Three True Outcomes players is highly dependent on their offense. When the offense is going well, and for most of these types of players it is usually going well, then the player can be incredibly valuable. However, when Chris Carter or Adam LaRoche cannot get things going offensively, he provides little value to the team. More players like Joc Pederson, Mike Trout, or even Bryce Harper could help change the stereotype of the lumbering, Three True Outcome slugger, but given the data over the past few years, that stereotype still holds true.