The Cardinals Offense Has Been Maddeningly Inconsistent

The Great Big Book of Sports Clichés dictates that I’m supposed to begin any article about inconsistencies by making a Jekyll and Hyde reference. And while I do love the tar out of classic movie monsters (Wolf Man for life!), instead, I’m going to start this article by being very frank with you. This is a very hard team to write about right now. The weekend debacle against Atlanta concluded a 50-game stretch that saw them go 22-28. Then earlier this week, John Mozeliak made some very ill-advised and embarrassing public comments, singling out Dexter Fowler for a lack of hustle. I’m not here to re-litigate the drama from (*checks calendar*) three days ago. But it does provide context for the Malaise Forever creep that’s going on amongst even the most ardent fans, especially when coupled with the on-field success starting to circle the porcelain for the third year in a row. The series in Arizona was nice, along with last night’s game, but we’ve seen them play well and slide back again and again. You’ll forgive me if I need to see more before my enthusiasm is fully restored.

This malaise manifests itself for me in this article. Originally, I wanted to continue my series from last week and take a look at what relievers tend to cost via trade in the month of July. As of today, what’s the point? There are legitimate bullpen issues to address, but:

• The bullpen has started to correct itself of late (I remain adamant that they need help from the left side)
• Adding to the bullpen is rarely as impactful as this team would need. Tacking a generic reliever or two on to a 22-28-in-their-last-50-Cardinals team is like adding a splash of Four Roses Single Barrel to a flat Vess Cola. It might make things better but you’re still drinking flat, terrible Vess, and wasting some of your good bourbon. And what the hell are you doing mixing good bourbon (trade acquisition relievers, enhanced playoff hopes) with flat, generic cola (a mediocre team) in the first place?

Instead of writing about any potential deals, even minor deals for bullpen arms, I’m going to table that for now until it looks like that info would be pertinent (e.g. if they start winning again and playing better than a .440 clip). Instead, today I’m going to discuss the maddening inconsistency of the offense this season.

One way we can identify inconsistency is simply year to year collapse when improvement, or at least treading water, is expected. Last season’s offense was far from perfect, but it was viable and generally solid. Here are the team’s ranks in various important offensive categories, comparing last season to this season through Monday’s games. All stats are for non-pitchers.

That’s not a bad 2017 profile. Their plate discipline was great, their overall production was above average (top 10 even, via wRC+), and they made plenty of loud contact. Isolated slugging lagged behind the other indicators, but it seemed like an easy fix. Add punch to the lineup and it should translate more of those other indicators into runs. In theory, adding some punch would prevent the offense from going stagnant as it did so frequently last season. With that in mind, underperformers like Randal Grichuk, Stephen Piscotty, and Aledmys Diaz were swapped out, replaced with newcomers like Marcell Ozuna and Yairo Munoz. More playing time would be created for Jose Martinez, adding further thump to the order. It was a good plan. As you can see in the table above, it hasn’t worked out, with every category except Hard Hit % seeing some regression.

Now let’s add a few simple slope graphs to show where this plan has gone awry. Or rather, let’s see where inconsistent year to year performances have cost them. In this case, I’m showing only wRC+, going from a 2017 performer to their 2018 performance. As you’ll see, I’ve paired up a few players with their 2018 replacements. Some were easy- Eric Fryer to Francisco Peña, for instance, or Aledmys Diaz to Yairo Muñoz. Since 2017 Grichuk got more playing time than Piscotty, I linked him up with 2018 Ozuna. Then I linked 2018 Bader up to 2017 Piscotty. Finally, 2017 Bader is linked up with 2018 Tyler O’Neill. It’s not perfect but it’s not far off. I’m also showing production by position to the left. The 2018 data is through Monday’s games.

I’ve highlighted the biggest declines, and they are exactly what you’d expect. This graph illustrates two things you might not have known. First, it’s easy to know that Pham, Gyorko, Wong, and Fowler have declined, or that second base and rightfield have declined, but our slope graph here shows just how big the fall has been. Pardon the logistical impossibilities for a scond, but imagine if last year’s team had given Gyorko’s at-bats to 2017 Bader, Pham’s to 2017 Diaz, and Martinez’s at-bats to 2017 Piscotty. Those are comparable drops to what we’ve seen this year from Pham, Wong, and Fowler, respectively.

The second thing you can see in the slope graph that you might not have known is that pinch-hitting production has cratered. I’m sure there’s a ton of variance from year to year in team pinch-hitting production, so it’s impossible to draw any meaningful predictive conclusions. But as for descriptive conclusions, keep in mind that these are typically high-leverage at-bats. Falling from a 118 wRC+ to 54 is more significant than you’d think. It’s like having a bench full of Adam Wainwrights to deploy in some of the most critical game situations for half of a season.

Inconsistency can show up in another way I’d like to discuss. I’ve mentioned it a few times over the last few weeks. Specifically, they can’t seem to get multiple hitters hot at the same time, or even producing consistently at the same time. Pham and DeJong were hot to start the year, but the impact was short-circuited by terrible starts from Carpenter, Ozuna, and Fowler. Eventually, Carpenter joined in the fun, but Ozuna and Fowler were still struggling, and Gyorko’s warm start had cooled considerably. Then DeJong and Molina got hurt, replaced with lesser hitters. Ozuna and Carpenter were engines for a time, but Pham went in the tank along with Gyorko. Yadi returned and hit for some authority, pairing him with Carpenter’s offensive outburst. But now Ozuna’s wonky again, DeJong is still injured, Bader has gone from viable to poor, and Gyorko still can’t get going against anyone other than the Padres.

Let’s dive deeper into this using 15-game rolling wOBA for the 14 main contributors thus far. That counts Francisco Peña, Tyler O’Neill, Greg Garcia, Yairo Muñoz, and Jedd Gyorko as the bench guys, plus the seven regulars and the Bader/Fowler platoon. What I’m about to do gets a little tricky and I don’t recommend it as totally sound analysis. But it’s a starting place. The top 10th percentile wOBA among non-pitchers for this season is .384 and above. The bottom 10th percentile wOBA for the season is .284 and below. I’ve collected every date in which a Cardinal hitter (of the 14 main contributors) were either .384 or above, or .284 and below, over their most recent 15-game stretch. Then I’ve accumulated how many the Cardinals had of each by date- hot players and cold players by date.

I mentioned that it wasn’t sound analysis but it’s a start. I took the team’s 15-game rolling average runs scored, plus the number of players above .384 and the number of players below .284 wOBA, and ran a quick multiple linear regression on it. Those data points in total predict 15-game rolling average runs scored with an r-squared of .31. It’s not great or even good, but it’s not bad for some fairly arbitrary data.

At any rate, once I had the number of hot (.384+ 15-game wOBA) and cold (.284 15-game wOBA) Cardinals by date, I could see how much of a surplus or deficit the team was running in positive production by date. In other words, I subtracted the number of cold players from the number of hot players after each game. Here’s how that looks. It’s ugly.

They peaked this season with 4 players running a 15-game wOBA of .384 or better at the same time on June 21, and again from June 23-26 (Martinez, Molina, Ozuna, Carpenter). From June 8th to June 26th, that quartet ran a .384 or better wOBA. That’s great, except the Cardinals also had some combination of Wong, Gyorko, Fowler, Pham, and even Peña rolling in the .284 or less wOBA territory for their previous 15 games. Despite the most players all year producing in the 10th percentile or better, the effort was sunk by a deficit for number of players in the lowest 10th percentile.

There have been 18 dates all year where they weren’t carrying a deficit, and only six of those were an actual surplus. In the other 12, they were at zero (the same number of hot players as cold players). The last surplus date happened on April 27th. Since the end of game #15 on April 14th- when 15-game rolling data was available- through Monday, the Cardinals played 68 games. In 47 of those, they had two or fewer “hot” players- players with a .384 or better rolling 15-game wOBA.

As for the number of days in which players have been “hot”, Wong has a single date in which his 15-game rolling wOBA cracked the 10th percentile. Fowler has zero, as do O’Neill and Peña (although it’s more forgiving for those last two). Muñoz has three, Gyorko has nine (none since May 15th), Bader has five, and DeJong only had two before hitting the DL (although he was surely on his way to many more days in the positive). Tommy Pham has 24 (tied for second most on the team), but none since May 20th. Carpenter also has 24 but didn’t have any until May 20th. Ozuna’s in fourth place with 20, but didn’t have any until June 3rd.

Martinez leads the team with 33. And to re-emphasize the point, that’s out of 68 games. Martinez has been hot half the time, then the next best players (Carpenter, Ozuna, Pham, Yadi) are hot about a third of the time. And it becomes a goat rodeo after that. You can see why there isn’t much overlap amongst hitters performing well at the same time. There aren’t enough hot streaks to go around to build up any sort of offensive consistency.

Now, in full disclosure, I can’t tell you what’s normal in this realm. I don’t have context for other offenses this year, or past years, or even the previous year’s Cardinals. That would require a much bigger study that would eat up a lot more time than I have available. What I can offer is league-wide top and bottom wOBA as the baseline. And using that, I think it’s safe to say that it’s really bad to have 15-game stretches where you have more players running up crappy wOBA (less than .284) than players running up top-shelf wOBA (over .384), especially when the deficit is two players or more who are cold. Hell, as recently as July 1st and 2nd, they had a season high deficit of five players. Five more players were cold than hot.

The good news is that Molina is back, which is already helping, and DeJong has rejoined the team after a brief, successful minor league rehab. Those are stabilizing forces, and they should at least smooth out some of these deep deficits they’ve seen. Keeping Carpenter hot and getting either (or both! why not both?) of Pham and Ozuna hot again would, theoretically, open up a lot of possibilities for this offense. As it is, Pham is showing signs of life. Unfortunately for this season to date, that type of consistency has been too much to ask. If it keeps being too much to ask, we’re all going to need something a lot stiffer than flat Vess and a splash of bourbon to get us through the season.