Three Thoughts About Jose Abreu

When Astros’ fans talk about the team’s offense, the conversation usually comes around to first baseman Jose Abreu. And it’s not just because he has the worst offensive stats among starters on the team. It’s also the fact that the Astros hoped he would upgrade the offense when they signed him to a three-year contract.

Abreu is 52 on both wRC+ and OPS+. That means his overall offense currently is 48% worse than the league average. This is a horrific start to the season for Abreu. The win probability Run Expectancy statistic indicates that Abreu’s performance has cost the Astros almost 1 win (.76 REW) compared to the league average hitter. Despite the bad start, the projection systems still expect Abreu to rebound in the remainder of the season. ZIP expects Abreu to perform at a 111 wRC+ and OPS+ for the rest of the season. Is this realistic? Well, it would require a rapid reversal in Abreu’s current performance.

I have three thoughts on Abreu’s situation.

1. Lineup Construction and Batting Order.

The Astros’ batting order frequently places Abreu at the No. 4 position, which is normally reserved for one of the two best hitters on the team. (Tango’s The Book suggests that the two best hitters should be batted at 2d and 4th slots.) In that spot, Abreu separates lefthanded bats Yordan Alvarez and Kyle Tucker. I have previously argued that Alvarez and Tucker can bat back-to-back because they are both unusually good at hitting LH pitching. The Astros have batted Abreu No. 5 at times. Dusty Baker has said this choice may depend on the number of LH pitchers on the opposing team’s roster. (See Chandler Rome’s tweet providing Baker’s explanation.)

Should the Astros drop Abreu lower in the batting order? When batting orders are within a normal range, I usually won’t argue about batting order choices—because typical batting order changes have minimal impact on the end-of-season runs scored. But, in this case, I would answer “yes,” because it is atypical to bat a 51 OPS+ batter in the clean-up position (4 and 5). Moving Abreu from 4 to 5 is an improvement. But I would prefer the 6 or 7 slot, if not lower.

In order to test this question, I have used the Baseball Musings Lineup Analysis tool (Morong, Arneson, and Armbrust). The tool uses OBP and SLG inputs for each starting player to develop the best batting orders (in terms of expected runs per game). The tool contains algorithms that simulate run-scoring opportunities over the course of a season for each position in the lineup.

In order to test Abreu’s lineup position, I used his current OBP and SLG. And for most of the lineup, I used players’ current OBP and SLG. Because Jose Altuve has just started playing this season, I used the ZIPS ROS projection for his OBP and SLG. I also expect an improvement in Bregman’s batting line, and I used a compromise 50% current and 50% ZIPS ROS for his OBP and SLG. The players’ batting stats and the results of the lineup tool are shown here. The 13 best lineups are shown below.

Some conclusions from this exercise:

• In all of the 13 best lineups, Abreu bats 8th. In 24 of the top 30 lineups, Abreu bats 8th; in the remaining 6 of the top 30 lineups, he bats 9th.
• Altuve is the No. 2 hitter in most of the top 30 lineups. For most of the 30 best lineups, Alvarez bats 4th, and Tucker bats 5th. In some of the 30 best lineups, Alvarez is No. 4 but Tucker bats ahead of him.
• The run-scoring difference between the best and worst of the 30 top lineups is insignificant. The difference is 3 - 4 runs per season. This confirms the view that the difference in typical batting order decisions is immaterial.
• What about the unreasonable lineups? For the 30 very worst batting orders, Abreu bats 4th in half the cases. In the other half of the very worst lineups, Abreu bats 2d. This would seem to confirm the notion that Abreu shouldn’t bat 4th and, in fact, should bat lower in the lineup.
• The difference between the 30 top lineups and 30 worst lineups is approximately 80 runs per year or 8 wins. This is significant. Now, keep in mind that the worst lineups have multiple choices, which are so unreasonable that it’s unlikely they would be included in a manager’s lineup card. But, even if batting Abreu at No. 2 or No. 4 only contributed 20% of the run differential in the worst lineup, it is still responsible for almost a 2 win reduction in the W/L record. This assumes that Abreu’s offense doesn’t improve and his batting order position remains the same all year. Hopefully, that doesn’t occur, which would mean that the impact on the final W/L record is overstated.
• I also ran the lineup analyzer with Abreu’s batting set similar to Bregman’s (50% current and 50% ZIPS ROS) as a sensitivity. The basic conclusion about moving down Abreu’s position remained similar. In some top 30 lineup cases, his position changed from 8 to 9. The difference between the top 30 lineups and the worst 30 lineups declined from 80 runs to 70 runs when Abreu is assumed to improve somewhat in the future.

I realize that the players are human beings and the batting order isn’t set by a computer. I suspect that Dusty Baker is hoping that Abreu’s 4 or 5 lineup spot enhances the player’s confidence. But I could also argue that dropping Abreu a few spots in the lineup would decrease the pressure he may feel, and could help him mentally. Maybe it would be too severe to drop Abreu to the 8 slot. But perhaps the 6 or 7 slot would be a move in the right direction in terms of lineup optimization, without sending a “lack of confidence” signal to the player.

2. Two Months Does Not A Season Make.

Yes, Jose Abreu’s batting results are bad. But these results have occurred in only 192 plate appearances. Approximately 72% of his season plate appearances remain. Some Astros fans have convinced themselves that his batting results can’t or won’t improve. I understand the frustration, but we must remember that baseball is replete with examples of reversals of fortune following several months of slumping performance. That is why I seldom conclude that a terrible slump shows that an aging batter is “toast” or has fallen off a cliff. Too many times, that impression turns out to be proven wrong.

At age 38, Willie McCovey posted an OPS+ of 82 in 250 PA. Was he done? The next year he posted an OPS+ of 132 in 548 PA. In 2016 Josh Reddick posted an OPS+ of 74 in 167 PA for the Dodgers. He then signed with Houston and posted a career-best OPS+ of 130. In 2020 Marcus Semien posted a first-half wRC+ of 61, but his second-half wRC+ was 128. At age 38, Rusty Staub’s OPS+ fell from the prior year’s 149 to 71 in 250 PA. How did he hit the next year at age 39? OPS+ of 123. Through age 29, Adam Dunn had 12 consecutive seasons with triple-digit OPS+. At age 30, he endured an OPS+ of 54 in 496 PA. He then proceeded to post three consecutive seasons with triple-digit OPS+. Or remember Hunter Pence, who posted an OPS+ of 59 in 330 PA at age 35, only to follow that season with an OPS+ of 128 for the Rangers in 2019. I think you get my point.

You can’t assume that 28% of the season dictates what will happen in the remainder of the season.

3. Slow Starts for the Free Agent First Basemen.

The Astros went into the off-season believing that they needed to acquire a first baseman. However, we knew then that the first base free agent market was relatively weak.

The top first baseman in the class was Anthony Rizzo, and the Astros reportedly showed interest in him. But that was probably a long shot because the Yankees seemed likely to re-sign Rizzo. And, in fact, the Yankees re-signed him early in the free-agent period. Based on past performance, Josh Bell and Jose Abreu were probably at the top end of the market after Rizzo was signed. Based on media reports, the Astros outbid Cleveland for Abreu. And Cleveland turned around and signed Bell to a 2-year contract. For whatever reason, most of the free-agent first basemen (other than Rizzo) have gotten off to a slow start through the first couple of months. Abreu’s OPS+ is the third worst of the group so far, but all of them are sub-par when you consider that first base is an offense-first- position— which should produce results around 110 OPS+.

Free Agent First Basemen (Current OPS+)

Bell 95 OPS+, 0 WAR

Myers 45 OPS+, -0.5 WAR

Voit 53 OPS+, -0.2 WAR

Hosmer 67 OPS+, -0.4 WAR

Mervis 51 OPS+, -0.3 WAR

Mancini 89 OPS+, -0.9 WAR

Abreu 52 OPS+, -1.1 WAR

C.J. Cron 80 OPS+, -0.2 WAR

Y. Gurriel 80 OPS+, -0.3 WAR

Santana 94 OPS+, +0.8 WAR

Rizzo 153 OPS+, 1.7 WAR

The Yankees’ signing of Rizzo has so far been spectacular since he sports an OPS+ 53 percent higher than average. At this point, the Pirates’ signing of Carlos Santana for $6 million appears to be the best deal among the remaining group. But it’s too early to really judge the off-season acquisitions. We can save that for a future day when more information is available.