clock menu more-arrow no yes

Filed under:

Of Black Holes and Regression Towards the Mean

New, comments

Will the Diamondbacks ever find consistency on offense this year ?

NASA’s WISE Telescope Reveals Millions Of Black Holes Photo by NASA/ESAvia Getty Images

Much has been written this year lamenting the lack of consistency displayed by the Diamondbacks offense . One would think that the roller coaster ride could not get any more extreme than the historical lows of May sandwiched in between the highs of April and the first few weeks of June.

However what we just witnessed the previous 48 hours perhaps is a 2018 microcosm : 20 runs on 18 hits including 5 homers on Saturday night, followed by a 16 inning 4-3 loss in which the team was held scoreless for the final 12 innings all while going 3 for 14 with RISP.

What is the cause of this inconsistency ? Well one major reason is the sheer number of “Black Holes” that have collectively gotten a significant number at bats this year. Updating a report, I posted in the Round Table discussion, the DBacks have 6 players with minimum 65 PA, OPS below .600 and BA below .200.

  • No other team in all of MLB has that many. D Backs top the list.
  • The Brewers have 3.
  • All the other contending teams in the NL have NONE. ZERO.

Below is a table view of these 6 players and what they have done, (or not done I guess) year to date. These 6 players have garnered 26% of the position player at bats, while putting up a .515 OPS and triple slash of .177/.261/.255


Alex Avila 126 .148 .270 .269 .538 42 4 17 54
Chris Owings 244 .189 .254 .279 .533 39 3 18 64
Deven Marrero 85 .167 .224 .205 .429 13 0 6 23
Jarrod Dyson 237 .189 .282 .257 .539 43 2 27 34
Jeff Mathis 115 .182 .278 .242 .521 38 0 14 34
Steven Souza 66 .148 .212 .213 .425 12 1 5 19
Black holes 873 .177 .261 .255 .515 36 10 87 228
All others 2401 .258 .335 .466 .800 106 91 225 534
Total 3274 .236 .315 .409 .724 87 101 312 762

On any given night, 2 or 3 of these players have been in the lineup, sometimes 4 or 5 of them at the same time ! Simply put, on the nights when your hitting stars such as Goldy or Pollock or Peralta don’t carry the team, which they can’t be expected to do every night, there have been too many automatic outs from these 6 guys. Not only have they not picked up the slack, they have dragged it down further.

Now we know that 4 of the 6 guys on this list are here primarily for their gloves, not their bats. But there are thresholds. No matter how good defense you play, or how well you run the bases, you have to reach at least a minimum level of offensive production to be a net positive player. Here is another view that illustrates this more clearly (I hope) :


Name PA Rbat Rbaser Rdp Rfield Rpos RAA WAA Rrep WAR
Name PA Rbat Rbaser Rdp Rfield Rpos RAA WAA Rrep WAR
Avila 126 -9 0 0 2 2 -5 -0.6 4 -0.2
Owings 244 -19 1 1 8 -1 -10 -1.2 8 -0.5
Marrero 85 -10 1 0 3 1 -6 -0.7 3 -0.4
Dyson 237 -17 3 0 9 0 -5 -0.7 8 0.1
Mathis 115 -8 0 0 4 2 -2 -0.3 4 0.1
Souza 66 -8 0 0 -3 -1 -11 -1.2 2 -1
TTL 873 -71 5 1 23 3 -39 -4.7 29 -1.9
others 2590 0 0 3 27 31 61 5.3 75 12.6
Total 3463 -71 5 4 50 34 22 0.6 104 10.7

So what do we see here ? Well the linear weights batting runs above average (also know as wRAA, which is the same as rBAT column above) are clearly very very much negative. There is some offset with positive defense and base running for most of these players , which mitigates some of the offensive drain. But not nearly enough. They are ALL well below average, (see WAA or Wins Above Average), and only Dyson and Mathis even approach replacement level, (just barely at 0.1)

Remember replacement level players ( or 0 WAR Players ) are well below MLB average players. ( Avg = Approx 2 WAR for a full season) And in the case of the other 4, they are well below replacement level. Chris Owings for example, has played great defense this year. There is no denying that. We can see it both on T.V. or at the ballpark, (and in the numbers as well.) But his hitting has been just SO strongly negative, (-19 rBAT) no amount of defense and base running is going to make up for it.

Also very important to note: You can see that the team has -71 rBAT in total, and that essentially all of the -71 has come from these 6 players. The rest of the team, including the pitchers, has been league average in aggregate, ( or 0 rBAT)


Only one of the players on this list is a hold over, the aforementioned Owings. The other 5 were all brought in by Mike Hazen. Jeff Mathis was signed prior to 2017 to catch Zack Greinke, and the other 4 were brought in for 2018. So these are “his guys”.

Jarrod Dyson while never a “good” hitter, was never this bad. He was actually pretty consistent around his career avg 85 OPS+. Deven Marrero was always strictly a glove guy. Alex Avila (105 OPS+ thru 2017) and Steven Souza Jr. (106 OPS+ thru 17) were brought in because of the combination of their league average to a little above league avg batting track records, PLUS their defense was projected to add value as well. Yet here we are with their numbers so low that it defies both logic and belief.

So did Mike Hazen make bad evaluations and bring in a bunch of stiffs, or over do it with the emphasis on run prevention at the expense of sub replacement level offense ? OR has there been an inordinate amount of bad luck ? For the sake of simplicity , the below tables only show OPS. Ranked in order of projected OPS


Player Career thru 17 2017 Proj YTD YTD vs. Proj
Player Career thru 17 2017 Proj YTD YTD vs. Proj
Souza .751 .810 .781 .425 -.356
Avila .752 .834 .741 .538 -.203
Owings .685 .741 .713 .533 -.180
Dyson .667 .674 .682 .539 -.143
Marrero .568 .593 .592 .429 -.163
Mathis .565 .600 .584 .521 -.063

The depth of despair in these numbers is palpable. As repeated often, not all of these guys were considered to be decent offensive players in the first place. Guys like Jeff Mathis and Marrero seriously under performing their projections don’t really shock anyone. However Dyson and Owings being over .140 and .180 OPS points below their projections has been a surprise. And then we come to Souza and Avila. The sample sizes are still small. Souza 66 PA, Avila 126 PA. But the gaps between their projections and their track record, and their out put so far, are off the charts bad.

One might say that you could look at each of Souza and Avila over last 2 months of 2017 and figure maybe something was up. Souza had a .880 OPS on July 31st last year, but hit just .650 the rest of the way. As late as July 2nd last year Avila had over a 1.000 OPS and a .310 BA in over 200 PA. From July 4th on he had a .634 OPS and a .210 B.A. But thats really parsing the game logs and not a good way to evaluate talent. I don’t think we can fault Hazen here. These were not bad evaluations.


All the above said, and as bad as these guys have all been, individually and collectively, there is hope. That hope lies in regression TOWARDS the mean. I strongly suggest reading through that article linked. Here is a key section:

“Any sample of PA contain potentially useful information. Maybe he’s hurt, maybe he’s aging poorly, maybe the league learned to exploit a weakness. Maybe his true talent has changed. But when we are asked to assess this player, the previous five seasons carry a lot of weight. We don’t just forget about them because our player had a bad April. So to forecast his future performance, we need to consider RTM. It’s more likely that he will perform close to his career average (or some weighted version of it) than the sample of plate appearances immediately preceding the question.”

And therein likes our hope for this group of players, and accordingly hope for some semblance of consistency for the entire team to emerge. Below are their Rest of Season Projections . Yes, I can hear the groans. They all just SERIOUSLY missed their projected performance levels for the first half. (Or the projections missed them !) And some of them probably will again in the 2nd half as well. In fact, each of them has seen their ROS projection lowered from their original pre season projections, precisely because of how bad they’ve been. But the odds strongly favor of at least 2 or 3 of these guys to regress towards the mean. And except for Mathis they are all projected to improve by over 100 OPS points for the ROS, and in Souza’s case over 300 points !


Player YTD ROS Proj ROS vs YTD
Player YTD ROS Proj ROS vs YTD
Souza .425 .769 .344
Avila .538 .693 .155
Owings .533 .667 .134
Dyson .539 .661 .122
Marrero .429 .568 .139
Mathis .521 .567 .046

Final words on this: Obviously the “key guy” on this team is Goldy, followed by core players Pollock, Peralta, Lamb, Marte and even Ahmed. They need to all stay healthy and continue to hit. But as mentioned , 26% of the team’s PA’s have gone to the 6 “Black Holes” I highlighted here. If 3 of those 6 can improve as projected, and the core group can continue to balance each other out, the D Backs should be able to find some consistency going forward.

Fingers Crossed