28 May 2015

The Orioles Were Right Not to Re-Sign Nelson Cruz

This is a guest post by Luis Torres. Read more of his work at Taking Back Baseball. You can also follow him on Twitter.

During the first game of the Mariners/Orioles series at Camden Yards, I was perusing my Twitter feed and came across a conversation between Jon Shepherd and Matt Kremnitzer. They were discussing the frustration that O's fans have had over their team's inability to re-sign Nelson Cruz during the offseason. Their frustration with the organization's lack of major acquisitions this past offseason reached a boiling point when Cruz hit a home run off of Miguel González that night. Meanwhile, Orioles' corner outfielders have all been performing at replacement level this season.

It's easy to overreact when you see that one of the best offensive contributors for your team last season is hitting dingers against that very same team this season. With the AL East being as close and competitive as it has been, one could assume that Cruz could've been a difference maker. However, you can't draw such a conclusion six weeks into the season, and you certainly can't draw conclusions on a four-year contract seven weeks into year one.

Contrary to popular belief, trades and free agent signings are to be evaluated when they are announced. Not during, and not after the contract is up, but before the player's first game is even played. In baseball anything can happen, including good results from bad decisions. The only fair, logical way to evaluate acquisitions is to go by the information that you know at the time. Let's take a look at what we knew about Nelson Cruz when the Mariners signed him.

The Mariners are paying $58 million for the age 34-37 seasons of a player who is a slow, poor outfielder, and is likely to age poorly due to his lack of athleticism. He's basically a DH who is a low OBP, high SLG hitter. I won't go into too much detail as to why I didn't like this deal for the Mariners, suffice it to say that their general manager, Jack Zduriencik, keeps paying for power when his team has finished last in the AL in OBP in four of the last five seasons and were second to last in the other.

By being smarter than Zduriencik, the Orioles avoided paying $58 million to an aging slugger and got a compensation pick in the process. That may be of little consolation to Orioles fans right now, but the thing is that Cruz is grossly overperforming and is going to come back down to earth. Remember him doing the same thing last year? Through May 2014, Cruz enjoyed a 27.9% HR/FB ratio and .324 BABIP, which fueled his outstanding .440 wOBA and 19 HR. Seeing how his career rates consisted of a 16.6% HR/FB and .302 BABIP going into that season, it was obviously unsustainable and proved to be so. From June 1 through the rest of the season, those rates dropped to a 15.9% HR/FB and .272 BABIP, resulting in a .331 wOBA and 20 HR. In other words, he turned back into Nelson Cruz. Not that Cruz playing like himself was a bad thing, especially for only $8 million. It was actually one of my favorite signings prior to the 2014 season.

Cruz is currently hitting .341/.398/.688 with 17 HR, which is good for a .458 wOBA and 2.2 WAR. He's leading the AL in AVG, SLG, HR, Total Bases, and wOBA. His OBP is second only to Miguel Cabrera. He's even better than he was at this point last year, and like last year, it's because of even higher and flukier batted ball stats. He currently has a whopping .376 BABIP and 30.9% HR/FB! This is all without any significant changes to his batted ball rates, too. For how much longer is it reasonable to believe that he's going to continue hitting almost one-third of his fly balls over the fence? Cruz is grossly outperforming even his 90th percentile PECOTA projections of a .351 wOBA by 107 points!

About all those home runs -- let's take a look at the pitchers who gave them up to Cruz. The following numbers are the pitchers' ZiPS projections going into this season.

Collin McHugh
at Houston
Miguel González
at Baltimore
Brett Anderson
at Oakland
Wandy Rodríguez
at Texas
Phil Hughes
Dan Otero
at Oakland
Tyler Clippard
at Oakland
Luke Gregerson
at Houston
Matt Shoemaker
at Los Angeles
David Huff
at Los Angeles
Brandon McCarthy
at Los Angeles
Ross Detwiler
Frank Garces
San Diego
Marco Estrada

There are a few good pitchers in there, but overall it's an underwhelming group.

At the time of the signing, Dan Szymborski, who is also an Orioles fan, tweeted out his disdain for the deal. In the tweet, Szymborski included the ZiPS projections for Cruz's years with the Mariners, assuming he'd be their primary DH. Remember that projections are not predictions. They are a measure of true talent level. Here's an excerpt from that table:


While I'm not a fan of the generic $/WAR metrics out there because I believe that they're too oversimplified, $58 million and a first round pick for 4.3 WAR is a pretty bad deal. Furthermore, the ZiPS model may be overestimating the number of games that Cruz will play. He's been a bit injury prone in his career, having played over 130 games in a season only twice. Obviously, getting older isn't going to make that any better. Even with his poor defense, he'd still be an upgrade over the poor corner outfielders that the Orioles have been trotting out there, but it'd come at a high price. By the time his contract enters its third year, Cruz is likely to be just as bad as those guys anyway.

OK, those were the numbers, so now let's take a look at the scouting information. Mark Anderson wrote up a scouting report for Baseball Prospectus this past October. I won't go over it in detail, but I will show the scouting grades. For those of you who are unfamiliar, these grades come from the 20-80 scouting scale.


Nelson Cruz is currently playing like an 80 hitter with 80 power. Those players are incredibly rare. The only player in today's game who fits that description is Miguel Cabrera. There's only a handful of players who fit that description in the past 20 years: Albert Pujols, Manny Ramírez, Barry Bonds, Frank Thomas, and...that's it. It truly is a uniquely special talent to be able to combine an 80 hit tool with 80 power. Is it really reasonable to believe that Cruz has turned himself into this class of player over the course of one offseason? I doubt that these grades come as a surprise to anybody who has seen Cruz play. From a qualitative perspective, he has big power but is only a mediocre hitter, and his speed and fielding have gotten to the point where he should be at DH as much as possible.

I hate to have to discuss this, but because somebody is going to bring it up, I don't believe that steroids have turned Cruz into a completely different player. We know for a fact that he's taken them in the past, and if you check his Fangraphs page, you'll see that he's never hit like this before. In fact, his success on the Rangers is better explained through his hard work and openness to instruction. Out of all the admitted steroid users, none of them ever turned from average to Miguel Cabrera overnight.

It is possible that Cruz has found a way to tap into some more power, though at the age of 34 that's highly unlikely. I can, however, absolutely guarantee you that Cruz didn't suddenly become an 80 hitter. Going from a 50 to an 80 hitter requires an enormous increase in the skills required to hit for average. It just doesn't happen outside of young prospects that are still developing.

Matt Kremnitzer did a nice job of explaining the Orioles' corner outfield problem recently. Even a Nelson Cruz performing at his normal levels would be an upgrade over what the team has been trotting out there, but not as much as you'd think. Cruz gives back a good chunk of his offensive gains as a result of his poor defense. He already has a -5 DRS and -7.6 UZR. His poor DRS adds up to a -0.8 dWAR, which is pretty bad for a full season, let alone seven weeks into it.

It's not like the Orioles don't have their own player who is overperforming. I won't go into the same detail that Matt went into when he wrote about Jimmy Paredes, who has taken over for Cruz this season as the team's primary DH for the most part. He currently has a .394 BABIP and 23.1% HR/FB. Those numbers will certainly regress, but thanks to his work with Robinson Canó during the offseason and changes in his approach, it's possible that his true talent level is better than the 73 wRC+ that ZiPS projected going into this season. However, I'm sure that it's nothing close to his current 155 wRC+.

Regardless, depending on whether you're going by Fangraphs or Baseball-Reference, Cruz has only been approximately half a win to one win better than Paredes so far. The reason why that difference isn't larger is the defense. Although Paredes certainly won't be winning any Gold Glove awards, his defense is a full grade better than Cruz's. It's also worth noting that Paredes is making the league minimum. Cruz, on the other hand, is making $12.5 million this season.

So for those of you out there who are still mad at the Orioles for not signing back Cruz, given all the information that I laid out here, let me ask you: Do you really think that Cruz is going to continue hitting like this for much longer? Would you risk four years and $58 million on an aging slugger to find out? Or would you rather just spend the money elsewhere and take the draft pick instead?


Jon Shepherd said...

One thing I would like to emphasize is that this team is obviously operating at maximum payroll capacity (regardless whether you believe they can spend considerably more). At that point, surplus value is highly important whether that surplus comes from cost controlled players like Caleb Joseph or potentially overlooked talent like Wei Yin Chen. Paying elite value for a player is simply not viable for the Orioles. Cruz at 15 MM was thought at the time to be fair immediate value, but very likely to be an albatross of a deal toward the end. That situation of negative value and a salary block would have been almost disastrous in 2017 and 2018.

Anyway...solid article.

Luis Torres said...

All good points, Jon. Thank you!

Matt Perez said...

Parades has played an entire 52 innings in the field this year and has since been moved to the DH due to his poor defense. When Steve Pearce starts at second base instead of you then you know you're bad. His UZR is almost definitely a function of small sample size and would be expected to plummet over time. The only difference defensively is that Cruz is in the field when he shouldn't be while Parades is at DH like he should be probably meaning that Cruz is more valuable defensively than Parades. Ultimately, you can put Cruz at a position even if he'll be bad but you can't do that for Parades because he'll be terrible. I wouldn't just look at their "defense" score in this case.

It's only fair to judge whether the Orioles should have signed Cruz based on what we knew about him at the time. But we can use what we've learned about his performance to determine whether it was a fair deal. So far, his performance has been worth about $14 million meaning that he needs to be worth 3.7/44 plus a draft pick. Figure that's 6 more WAR?

It's still questionable whether he'll be worth that but it's ultimately more likely that he'll be worth his contract now than it was 50 games ago.

It's a good article. Just a few thoughts.

Jon Shepherd said...

A couple things. (1) From the Mariners point of view, you have to tack on another 10 million or so with their lost draft pick (I forget where it is located). (2) While looking at contract performance is a good way to judge whether or not a deal was good, it does not inform you whether the process that went into offering the deal was sound. The concept of a fair deal is something I would think is locked in a Winter of 2014 frame of reference unless we have good evidence that the Mariners knew something that no one else did. In other words, if I agree to help out a Nigerian prince in order to get 10 million dollars and wind up getting 10 million dollars, I would suggest that the deal was not fair given what we know about helping Nigerian princes out unless I actually had information meaningfully different than what most people know. Maybe this is all a semantics issue, but the process is what concerns me most.

luistorres16 said...

Matt: I believe what you're trying to say is that Cruz has more positional value than Paredes even though he fields his position poorly. I haven't seen Paredes play the field myself, so I can't say how he compares to Cruz defensively. I can tell you that Cruz currently has a -5 DRS and -7.6 UZR, which means that, regardless if you go by fWAR or bWAR, he's already more than negated the positional adjustment boost he gets for OF over DH for 2 months of the season. The way Cruz is hitting, it's no big deal, but once he turns back into himself, it will be. Even if Paredes is worse defensively than Cruz, you could still say the same thing about him, too. The thing about Paredes is that his bat won't play at DH once his performance regresses to the mean.

It's entirely possible that Cruz surpasses all our expectations and returns equal or better value for his contract. I would call that a case of bad process, good results. i.e. Zduriencik got lucky, which leads me into Jon's point.

Process is what concerns me most, too, and it's from that perspective that I criticize the Mariners and vindicate the Orioles. Of course, that assumes that I know everything that the Mariners knew at the time. When it comes to teams with great analytics departments, such as the Royals and Pirates, or GMs with great track records, such as Dave Dombrowski, I'm more concerned that they're going by private information that I don't have when they make deals with which I don't agree.

With Jack Zduriencik though...ehhhh. Given what we know about how he thinks about baseball, it's likely that all he did was look at Cruz's triple crown stats from 2014, and his signing does nothing to fix the Mariners OBP issues. Even with Cruz performing at his current level, they're still 2nd to last in the AL in OBP.

Thanks a lot guys for the good comments and feedback!

T. Berry said...

I like the front office's strategy of maximizing the number of quality players rather than overpaying for a few really good ones. Adam Jones and Manny Machado are the guys you keep on the payroll, no matter the cost, and build around for 5 to 10 years. The Orioles are at their best when employing two or three superstars with a mix of hard scrabble big leaguers.

Anonymous said...

There are a few logical inconsistencies in this post. First, I am not sure what ZIPS projections are worth. Even if take at face value, however, ZIPS projections are park-adjusted, so those projected numbers were in Safeco, one the most pitcher-friendly parks in the AL. Cruz's ZIPS projections in Camden Yards would have been significantly higher. Indeed, the vast majority of his home runs have been hit on the road. To be generous, you are using the wrong ZIPS projections.

At the end of the article, you switch your argument approach and compare Cruz to Jimmy Paredes. Before this season, however, Paredes was a waiver-wire backup third baseman with a career .750 OPS in the minors. If you say that Cruz's hot start should not be taken at face value because it wasn't foreseeable, then you cannot logically argue in the same sentence that the Orioles knew they could turn to Paredes to be DH. No one could have known that a mediocrity like Paredes would become the team's best hitter so far.

The discussion of Cruz's defensive liabilities is almost comical. I realize Cruz is not a good OF. Unlike batting stats, however, defensive stats like UZR are highly subjective. Arguing that Cruz costs his team nearly as much on defense as he gains for it on offense is subjective at best, laughable at worst. Besides, the Oriole corner outfielders can neither hit nor field. Even if one were to accept your argument at face value, which I would not do, at least Cruz can hit a whole lot better than the scrubs we are using in left or right field, none of whom can field either.

The stuff that you include from scouting reports makes even less sense. It is so jumbled that I do not even know where to start. Suffice it to say, Cruz is an elite hitter, and to answer your question, I would certainly take Cruz for his current contract. I would not have minded paying him $58M/3 yrs. I do not expect anything during the fourth year and see it as a throw-in. This team was built to win now, with 11 free agents at the end of the year. That will be harder to do without Cruz, as you are seeing already. Considering how we have been drafting, I would rather have Cruz than the draft pick.

Tim said...

I think you are overemphasizing park factors and the guy who creates the ZiPS projection agrees that signing Cruz would have been a poor decision to make given what the Orioles likely knew at the time.

I agree the Paredes comp feels like the column needed more inches and does not add much. A better argument would be comparing breakouts of similar waiver wire types to figure out a probability.

Defensive stats do a good job of describing what has happened similar to how RBIS describe what happened. Both have trouble describing what will happen. Cruz is one of worst outfielders whom teams allow to play the field regularly. I think at nest a Cruz in the field take a potentially great bat and makes it rather normal.

Finally...maybe a better question is would you have bet on Cruz for this year while having some understanding that the last three years of the contract mean little because this deal blows away many of your options in retaining your upcoming free agents or do you invest that 2015 money in one year of Chris Davis and Alejandro De Aza. I think in October or November that answer was clear. If Duquette had a crystal ball...the answer becomes slightly less clear.

Anonymous said...

Park factors matter a great deal. If you do not believe me, read Bill James' analysis way back when Freddy Lynn left the friendly confines of Fenway for Anaheim. I am surprised that you would even put that out there as an "argument." Do you seriously mean there is no difference (or little difference) between playing for the Yankees, Orioles or Rangers as opposed to Oakland, SD, or Seattle? ZIPS projections are pretty meaningless to begin with for the same reason that earnings projections are meaningless in evaluating stocks. Even if I were to take them at face value, Cruz should be more productive in Camden Yards than in Safeco.

I do not quite buy your third paragraph (about defense). If it were true, then why not just sign a bunch of .700 OPS guys who play great defense? That is not too far from the Rays' model, and it has its limits. I do understand what you are saying, but I would gladly take his offense even knowing his defensive limits.

My answer to the fourth paragraph is a resounding yes. I wanted them to get rid of Davis and thought of De Aza as a mediocre knucklehead at best. He had some talent, but he cannot field, often throws to the wrong base, and is an awful base runner. We are gong to lose most of our 11 FAs anyway, so I wanted acouple of good years from Cruz, which I thought we would get, and worry about 2016 at the end of this year. Angelos never spends enough anyway.

Anonymous said...

You may be surprised by James' more recent work about how way too much is made of park factors.

I do not think you understand third paragraph. I said they are good descriptive statistics but poor predictive statistics.