20 March 2014

Orioles' 2014 Offense Appears to Be as Unimpressive as Last Year's

by Daniel Moroz
Camden Depot and Camden Crazies

People seem to be optimistic about the Orioles' offense this year. Really optimistic. I've heard it on the radio and seen it on Twitter. Comments tend to be along the lines of "the offense was really good last year, and they added Nelson Cruz!" Being more detached from the sport than I used to be, I figured I was missing something.

2013 Team A: .255/.327/.420, .328 wOBA, 109 wRC+

2013 Team B: .257/.330/.409, .325 wOBA, 109 wRC+


An Offensive Celebration | Photo by Keith Allison
These teams finished 4th and 5th in Majors in wRC+ last year (excluding pitchers), behind the Red Sox, Tigers, and Cardinals.  Which one was the Orioles?


Trick question, they're the A's and Rays. The O's were 14th with a very average 101 wRC+. 


My guess is that because of the team starting out hitting better, Chris Davis' amazing season, the Major League leading 212 homers as a club, and park effects (the .324 wOBA makes them look about as good as the Rays), people overrated how well the 2013 team hit. Memory can be tricky like that, along with one's lying eyes (but man, there were some impressive home runs).


Davis was indeed great. Adam Jones was good. Danny Valencia and Steve Pearce added a bit off the bench. Then there were Manny Machado, J.J. Hardy, and Nate McLouth each around a 100 wRC+ with pretty much everyone else falling below that. Overall, the O's hit OK last year. Why would we expect a substantial improvement this year?


Matt Wieters posted a .302 wOBA in 2013 due largely to a low BABIP, but ZiPS has him rebounding to .321. Nick Markakis will hopefully be better than the career low .356 SLG and .304 wOBA he posted (ZiPS projects a .322). Manny Machado should improve some (assuming no after-effects from his injury). And Nelson Cruz's .340 wOBA (ZiPS) would be a vast improvement over the .304 the O's got from the DH position last year.


On the other hand, Chris Davis probably isn't likely to repeat what he did (.421 wOBA to .377 ZiPS). David Lough is projected to hit worse than Nate McLouth did last year (.296 vs. .323). Plus 300 PA of ~.365 wOBA off the bench (Valencia and Pearce) isn't expected. 


Throwing all the ZiPS* projections together with some guesses about playing time produces a wOBA in the neighborhood of .325, which is marginally better than 2013 -- the plusses and minuses mostly cancel out. And playing half their games in Camden Yards means that'll translate into a wRC+ that's above 100 again. A little above (maybe 102) -- which can be perfectly fine. If the Orioles are going to make the playoffs though, they'll need the pitching and defense to be sound, as it seems unlikely to me that the offense will carry them the whole way by itself.

* Steamer is slightly higher. Oliver is more positive still, projecting Chris Davis as the third best hitter in the Majors. PECOTA appears to lean the other way.

25 comments:

Bret said...

The Orioles becoming a good team was always contingent on Wieters (#5 pick, #1 MLB prospect, highest bonus in team history) becoming a good player. Last I checked, a .287 OBP isn't good. I don't disagree with most of what you are saying but I do think you are leaving out some things. Backup catcher was an absolute black hole last year - Teagarden and Snyder had 80 at bats and 12 hits combined. Urrutia and Reimold were very bad with nearly 200 at bats. I'm not so sure in an absolutely honest world where salary and prestige mean nothing if Clevenger shouldn't be starting against every RHP with Wieters only LHP/defensive replacement. Clevenger's minor league OBP is .373. Wieters has proven he can't hit left handed. But yes, the team can't get on base, Machado, Jones and Hardy view walks like the black plague and Lough and Cruz have major issues there also. I just don't think the team can be blamed for drafting a so-called franchise cornerstone, paying him a huge bonus, bending to his every need and having him stink.

Lou Proctor said...

I'm sorry, but I disagree with the premise of this article.

wRC+ is a very interesting statistic for trying to normalize and compare individual player performances. However, I question how valid it is in determining the offensive performance of a team.

A team's offense has one job: score runs. The Orioles were one of the best teams in MLB last year at doing so. Their 4.60 runs per game was good for 4th in the AL and 5th in MLB. How a team scores those runs is irrelevant and in the case of wRC+, seems weighted towards teams that produce runs a certain way (specifically, placing a heavy weight to OBP).

The Orioles, as we all know, are less than stellar at drawing walks and are relatively average at getting base knocks. But when they make contact, boy do they make contact. Despite the jokes about a "rally-killing HR", I don't think anyone would seriously take a BB over a HR.

Additionally, individual stats can't always be applied to team situations, because what makes for one type of good player doesn't necessarily make for a good team overall.

All in all, looking back at an offense that scored the 5th most runs in MLB in 2013 and saying it was "unimpressive" strikes me as being along the same lines as Keith Law's "there’s literally nothing that the Orioles can do to convince me that they are a good team" statement, even as the Orioles were tied for the second best record in the AL in September.

It's fine to look at underlying statistics to try and determine what's going to happen in the future. If you honestly believe the Orioles offense is going to be "unimpressive" in 2014, that's a perfectly valid analysis. But once you start trying to apply a single statistic to make a value judgment on a performance that already happened, a performance that was very good based on the only goal of an offense (score runs), you're stepping into very questionable territory.

TL;DR
1. wRC+ is an interesting statistic, but is not the be-all to evaluation.
2. Individual statistics can become even more flawed when applied to team situations.
3. For teams, statistical analysis for the future is interesting, but for the past, look at what actually happened. The Orioles' offense did score the 5th most runs in MLB, despite what other derived numbers say they should have scored.

Jon Shepherd said...

Interestingly enough, wRC+ is a team and league statistic as derived. It is then applied to players. That is how all context neutral statistics are derived. The actual concern is whether advanced metrics like these are well applied to players. For instance, Tillman being credited with 1.4 runs for each home run he gives up when in actuality, his pitching approach changes quite a bit with runners on base.

So, yeah, wRC+ has much better application to team performances than individual performances.

Daniel said...

Scoring more runs is good, but just looking at the total isn't that telling by itself. Is scoring 700 runs good? Well, for the A's (playing half their games in Oakland) it is, but for the Rockies (playing half their games in Colorado), it isn't.

You're also lumping AL and NL teams together when talking about MLB rank. The O's were 5th in MLB but tied for 4th in the AL (with all the other DH teams). Considering they play in a hitters' park, that's only OK.

Jon Shepherd said...

I think the argument is more that the runs will come just as they did which is a product of their talent but also their home field. In order to better compare offensive ability across teams, stadiums need consideration.

Daniel said...

Edit: Missed the 4th in AL part originally, but general point still applies - the O's scored ~40 more runs than the average AL team (but over 100 fewer than the best team) last year which is fine but not that impressive considering they play in a hitters' park.

Did get a chuckle out of "But once you start trying to apply a single statistic to make a value judgment on a performance that already happened", which is I guess fine if that statistic is runs.

Daniel said...

FYI - I didn't pick the title, so everyone yell at Jon about that. The only time I used the word "impressive" was in a positive way in reference to their home runs.

Bret said...

I think it is a little disingenuous to point out Valencia and Pearce but not mention Teagarden, Snyder, Urrutia, Reimold, Dickerson, Casilla, Morse etc. who ranged from horrendous to merely very bad. Clevenger should start against RHP, his minor league OBP is .373 and MW has proven he can't hit left handed. But yes, the team OBP is a continual issue. Jones/MM and Hardy view walks like cancer, that isn't going to change.

Jon Shepherd said...

Yeah...title was an editorial decision. The point I was trying to communicate is that there is a general perception of this team having an impressive offense, which requires some optimism. I used the word unimpressive to contrast with the perception.

Daniel said...

Thing I don't think is getting through;

a league average hitting team playing in Camden Yards is expected to score an above average number of runs,

a league average pitching staff (with average defense) playing in Camden Yards is expected to allow an above average number of runs.

The O's can have an average offense, a below average pitching staff, and a plus defense, and they'll (1) score an above average number of runs, and (2) give up a possibly average number of runs. This is what happened in 2013. Putting them in a neutral park might not change their record but we'd expect fewer runs scored (down to a more average number) and allowed.

Anonymous said...

If you are dinging the Orioles for playing in a "hitters" park, why did they have 720 hits at home and 740 hits away?

Matt Kremnitzer said...

I get why some wouldn't like the "unimpressive" label being used, but I think Daniel's points are pretty straightforward.

Something else not mentioned is that the Orioles were a relatively healthy ballclub last year. Davis, Jones, Markakis, Hardy, and Machado all played in at least 156 games; Wieters and McLouth played in 148 and 146, respectively. McLouth is gone, so left field will be manned by some combination of Lough, Cruz, Reimold, Pearce, etc. Machado should be able to work his way back slowly and will presumably be fine. But any injuries to one or two of those guys would cause some serious problems for a team that doesn't have much depth and doesn't really have any answers in Norfolk.

Matt Kremnitzer said...

2013 O's at home: .264/.317/.445.
2013 O's on the road: .256/.308/.418.

Anonymous said...

If I'm talking about a chef, I look at whether or not his food tastes good. If I were to start saying "sure it tastes good but what really matters is that he didn't follow the recipe and so I'm going to ignore the fact that he acheived his goal of something that tastes good" then that would be a total waste of my reader's time. It might be informative to say he didn't follow the standard recipe so it's less likely for him to recreate that taste next time, but saying that his delicious food wasn't very good is just foolish. The point here is that when you evaluate someone, you evaluate them on whether they met their goal and that's really it. When it's a chef, you evaluate their food based on taste. If their goal is to cook healthy then you can evaluate them on that too. What you doesn't make sense is to evaluate them on their ability to follow the same recipe as everyone else.

It's equally foolish to say that an offense that scored a ton of runs but didn't meet the bar set by advanced metrics wasn't good or wasn't impressive. You can say they won't score that many next year and give your evidence and start a discussion adn that's fine. The author can hide behind the title being an editorial decision but it's pretty clear from the beginning of the article that the title was based on the contents of the article. The 2013 Orioles offense did exactly what they wanted to do, which is to score runs. That's all there is to it as far as how good they were last year. The 2014 Orioles may repeat that success or they may regress to match the advanced metrics and that's a worthy discussion to have. I just think the introduction to this article caused the author to lose a ton of credibility for the rest of it..

As for the home/away thing, the Orioles ranked 4th in the majors in runs scored per game at home and 6th in the majors in runs scored per game on the road. I think either ranking is indicative of a very successful 2013 Orioles offense. Hopefully they repeat that success but it most certainly could go either way.

Jon Shepherd said...

Runs scored should be done within league in order to minimize impact of pitchers hitting. That said...runs probably shake out to an above average offense as opposed to exceptional. That really does not disagree with the article or the statements used.

Donovan said...

The thing (to me at least) that it seems as though commenters on this article are completely glossing over is that Camden Yards is not a neutral run-producing environment. Take a moment to consider what that actually means, instead of just glossing over it as being meaningless phraseology.

Camden Yards is an environment where it is less difficult (compared to other ballparks, on average) to get a desirable offensive outcome. wRC+ adjusts for this fact and makes it possible to compare offensive results as though they did all happen in completely neutral environments.

This is something that may be a tad counter-intuitive to conceptualize, but dismissing it entirely, in favor of the good ol' eye test does a disservice to what Daniel is trying to convey in this write-up.

As Matt mentioned, Daniel's points are very much straightforward. There is no wild conjecture here. To disagree with what he is saying is really to disagree with advanced statistics (specifically wRC+) themselves.

The chef analogy by Anon @2:11 could be adjusted to be workable: The chef who produces very good tasting food does it with the use of an oven that makes the food taste better on its own (an above-average flavor producing environment). While it may appear to your senses that the chef has accomplished some great feat, he was perhaps just a slightly above-averagely skilled chef cooking in an environment where "better" results are easier to produce. In measuring his true skill, or comparing his skill to that of other chefs, you would want to remove the "magic oven" from the equation, and you would see that even though his food tasted great, he is just a slightly above-averagely skilled chef.

Why, then, does true skill matter? Why not take the fact that the Orioles scored the 4th most runs in the AL as a victory and move on? Well, because of the run-producing environment they play in, their production, expressed in counting statistics such as runs, should be EVEN HIGHER than it was in 2013 in order to classify their offense as anything other than unimpressive (compared to the common perception of it).

Matt Kremnitzer said...

Now I'm just imagining Dan Duquette as a master chef. Pretty funny, actually.

Jon Shepherd said...

just a note...a commenter wished for his comments to be deleted. I deleted them, so it may seem choppy.

Jon Shepherd said...

I guess when it comes to chefs, it is easier to cook a great meal with a furnished kitchen than it is with one pot and a cord of wood.

Steve B. said...

To me the big problem last year (and it could be more of the same this year) is that the O's tended to score runs in bunches or not at all.

Lets say a team has a three game series. They score 10, 2, and 2 runs respectively. In runs scored space that equates to 4.7 runs scored/per game. Pretty dang good.

In the standings that probably left you 1-2. Sure this is a bit extreme but this is exactly how the O's hit last year and why they lost so many 1 run games that they had won the year before.

Jon Shepherd said...

I did something like that that might be useful to look at last year, but I did it a couple years back with 2012 data.

http://camdendepot.blogspot.com/2012/08/just-some-graphs-run-scoring-and-earning.html

Anonymous said...

I think the reason your these arguments are like two ships passing in the night is the park effects. The fact is that the Orioles are not competing against the other teams for their total runs scored, but rather their total runs scored against the Orioles. Each team playing in Camden Yards gets the park effect.

It is quite possible that the Orioles are 4th in total runs scored, but would be average if you corrected the league totals to show what they would be if all the other teams played half their games in Camden Yards.

Jon Shepherd said...

Right.

That is what the adjusted metrics try to accomplish.

duff77 said...

Gotta love all these new stats, but the only stat that counts is Runs. The Birds were 4th in the AL and 2nd in the East behind Boston last year. Sounds like a good offense to me. The problem is pitching, and nobody is going to convince me that Jimeniez and Yoon are enough to put us over the top. (Except, hopefully, the Birds themselves.)

Unknown said...

I agree, I don't see the offense being significantly better but about the same, and for the reasons given in the soundly based article. I specifically remember how bad the offense was down the stretch as opposed to earlier in the season. It seemed they couldn't buy a run when the needed to against the Rays, for example when facing them.