No, We Don’t Need to Worry About the Dodgers
On Tuesday night, the Dodgers were only a Gavin Lux home run away from falling to a .500 record. A .500 record isn’t generally cause for panic, but it would definitely have been a disappointment for the reigning world champions, a team that was expected to steamroll most of the rest of baseball this season. Just to match 2020’s regular season record, the Dodgers need to add another 24 consecutive wins to Monday night’s win over the Mariners.
So how worried should the third-place Dodgers be? Not very.
The Dodgers Aren’t Actually Playing Poorly
Okay, this header isn’t true if we engage in an ultra-literal reading, but in losing 15 of the last 21 games, the Dodgers have only been outscored by a total of seven runs in the aggregate. The team’s overall Pythagorean record puts them at a 94.5 win pace, below the preseason projections, but not alarmingly so. The bullpen had a 4.48 ERA over this stretch, and while there is a relationship between bullpen performance and Pythagorean performance, the relationship is fairly loose.
I went back through history to look at the Pythagorean performances of all teams that underperformed their expected record by at least two wins after 36 games. Over the rest of the seasons, those teams fell short of their Pythagorean records by about a tenth of a run on average. In other words, the discrepancy between expected record and actual record in the early season is mostly noise, as opposed to hiding something sinister about a team’s true abilities.
The Bullpen’s Outlook Hasn’t Changed Much…
Even if you could definitively point to the bullpen as the main cause of the 6-15 run, there’s still no reason to think the overall quality of the relief corps has changed significantly. The team’s bullpen outlook has gotten worse, with the group’s combined rest-of-season projected FIP going up by 0.03 runs (3.98 to 4.01) while the rest of the league has seen theirs drop by 0.12 runs. You’d rather be on the positive side of the ledger, but that only drops their overall ranking in our Depth Charts from sixth to eighth. And yes, those Depth Charts reflect the reduced playing time from Corey Knebel (down to just three projected innings) and Brusdar Graterol already.
…Nor Has That of the Rest of the Roster
While the projections for the rest of the team have declined slightly, it hasn’t been the equivalent of falling off a cliff. With one notable exception (see the next section), none of the hitters have seen their wRC+ projections drop by 10 points, nor have any members of the expected rotation seen their prognostications turn sour.
| Name | Preaseason wRC+ | Rest of Season wRC+ | Change |
|---|---|---|---|
| AJ Pollock | 98 | 102 | 4 |
| Chris Taylor | 101 | 105 | 4 |
| Zach McKinstry | 78 | 81 | 3 |
| Justin Turner | 120 | 120 | 0 |
| Max Muncy | 122 | 120 | -2 |
| Matt Beaty | 95 | 93 | -2 |
| Austin Barnes | 84 | 80 | -4 |
| Will Smith | 110 | 106 | -4 |
| Mookie Betts | 133 | 128 | -5 |
| Corey Seager | 132 | 124 | -8 |
| Edwin Ríos | 90 | 81 | -9 |
| Gavin Lux | 102 | 93 | -9 |
| Name | Preseason FIP | Rest of Season FIP | Change |
| Julio Urías | 4.15 | 3.90 | -0.25 |
| Trevor Bauer | 3.63 | 3.42 | -0.21 |
| Clayton Kershaw | 3.55 | 3.36 | -0.19 |
| David Price | 3.93 | 3.96 | 0.03 |
| Walker Buehler | 3.32 | 3.38 | 0.06 |
You’d rather improve slightly than decline, but the overall changes are fairly small. The changes in expectation have only dropped the rotation from second to third in our Depth Chart rankings. The position players have declined from the best overall preseason projection all the way to…second. Overall, we still project the Dodgers to be the best team in baseball over the rest of the season, just as we did a month ago. The only change is in the magnitude of that dominance.
The Roster Isn’t Radically Different
Naturally, injuries have the potential to turn projections upside down. One of the worst team mistakes ZiPS has ever churned out was for the 2014 Texas Rangers. Coming off a 91-win season, ZiPS projected the team to regress to 86 wins. Instead, the team went 67-95, dropping to last in the AL West. The largest difference between the projections and what actually happened didn’t come from the projections for the individual players, but who actually played. The pitching staff was decimated by injuries, basically losing the entire 2013 rotation for large parts of the season. The fill-ins were a parade of players with bleak forecasts: Joe Saunders, Scott Baker, pre-KBO resurgence Miles Mikolas, etc. Knowing who actually got the playing time and applying it to the preseason projections would have dropped the win projections under 70.
The Dodgers are not in that situation. Most of their key players are still expected to pretty much fill their anticipated roles this year. Losing Dustin May to injury is, of course, a Very Bad Thing, but in terms of the preseason outlook, he wasn’t one of the top 10 reasons the Dodgers were expected to win 100 games. There are other injuries as well, but the team’s IL isn’t bloated compared to that of other contenders. The Padres have lost Adrian Morejon, and there are questions about Dinelson Lamet‘s health. Atlanta lost Travis d’Arnaud, and Mike Soroka’s return has been pushed back to June. The Mets have been missing Carlos Carrasco and Seth Lugo. Eloy Jiménez and Luis Robert are missing from the White Sox.
The Hole Isn’t Deep Yet
The Dodgers are a game behind the Padres, but this race was always expected to be a dogfight, so that’s nothing new. Los Angeles stands three games behind the surprising San Francisco Giants, but the preseason expectations — 20 wins or more — provide a lot of room for error. If the Dodgers are actually that much better than the Giants, 126 games is far more runway than is needed.
So when should the Dodgers worry? To get an idea, I took the ZiPS in-season projections and projected how the divisional and playoff probabilities change with the assumption that teams continue winning at their current rate without any change in team strength. In other words, if the Giants continue to have a .611 winning percentage and the Dodgers keep underperforming and stay at .528, when does the team run out of margin for error? I checked the numbers after every 10 subsequent games.
| Team | Now | Game 46 | Game 56 | Game 66 | Game 76 | Game 86 | Game 96 | Game 106 | Game 116 | Game 126 | Game 136 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Giants | 0.9% | 1.6% | 2.7% | 4.9% | 8.2% | 13.7% | 22.1% | 34.6% | 51.2% | 71.0% | 89.3% |
| Padres | 44.6% | 46.3% | 48.0% | 49.3% | 50.1% | 50.2% | 48.3% | 44.0% | 36.1% | 23.9% | 9.9% |
| Dodgers | 54.5% | 52.1% | 49.2% | 45.9% | 41.7% | 36.1% | 29.6% | 21.5% | 12.7% | 5.0% | 0.8% |
| Team | Now | Game 46 | Game 56 | Game 66 | Game 76 | Game 86 | Game 96 | Game 106 | Game 116 | Game 126 | Game 136 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Giants | 12.4% | 17.4% | 23.6% | 32.0% | 42.2% | 54.2% | 67.0% | 79.5% | 90.1% | 97.0% | 99.6% |
| Padres | 90.1% | 89.7% | 89.2% | 88.6% | 87.7% | 86.8% | 85.6% | 84.9% | 84.4% | 85.0% | 87.1% |
| Dodgers | 93.1% | 91.7% | 89.4% | 87.5% | 84.3% | 80.1% | 74.9% | 68.6% | 61.3% | 52.9% | 44.5% |
Based on our current best estimate of the relative strengths of the NL West teams, the deficit doesn’t become concerning for the Dodgers until the end of June or so, leaving them plenty of time to improve the team at the trade deadline. However, an unpleasant surprise, such as a serious injury to another key contributor, could move that date earlier on the calendar.
The Dodgers faced a similar issue in 2018, to the extent that they actually briefly had a losing record after D-Day. They eventually won the division, but it was a close affair, requiring a 163rd game against the Rockies to grab the divisional crown. The Dodgers have an extra month to get off the schneid this time around. Until things turn in a more negative direction, the 2021 Dodgers are less the Titanic and more your average airline flight running into a bit of light turbulence.
Dan Szymborski is a senior writer for FanGraphs and the developer of the ZiPS projection system. He was a writer for ESPN.com from 2010-2018, a regular guest on a number of radio shows and podcasts, and a voting BBWAA member. He also maintains a terrible Twitter account at @DSzymborski.
You’re the man Dan. And on the plus side, Bellinger will return to help on both sides of the ball.
Well, the likely outcome is still that they win the division, just like they do every year. They are only three games back, and that’s pretty close.
But there were a couple other years in recent memory where they Dodgers cut it pretty close (2018 being the closest possible situation). I like that Dan ran the “when should you start to worry” question, because it’s important to remember that even if it hasn’t happened yet, at some point the banked losses are going to make up a large enough proportion of the schedule that you should worry. If the winning percentages are the same on June 12th as they are on May 12th, at that point you should probably start worrying a little bit (Dan says late June, but if they still have the same winning percentage by June 12th that means something else has likely gone wrong that would affect the play after that).
I don’t see people making this mistake so often anymore, but in previous years people would make arguments that the Dodgers shouldn’t bother upgrading because their team was already a lock to win the division (as measured by pre-season projections). But the error bars around projected records in the preseason (and early in the season) are far greater than we want to admit.
I’d say a big part of why the situation hasn’t changed all that much is that the deficit is pretty small, especially to the team most likely to keep up with them ROS. You can really see that in how it takes until game 116 for the Giants to be the favorite to win the division.
But as far as upgrading goes, this is actually pretty illustrative of why it is low impact: Thusfar the Dodgers are playing to something like their 10th or 20th percentile outcome by projection systems, and they’re still fairly sizable favorites. Similar in 2018: The end result was close, but it involved both a lot of way below-expectation outcomes (from Bellinger’s offense to Seager’s missing basically the whole season) plus an opponent that played way above expectation for the entire year (the Rockies) not just a month or two.
This is pretty close to the worst outcome over 36 games that projection systems contained inside their range of reasonable outcomes, and they’re still the favorites to win the division and huge favorites to make the playoffs.
Thanks, Dan! I wonder if ZiPS is still as bullish on Lux (seems not). He’s played a lot better over the last week, so I’m hoping he’ll click at some point…
It’s probably user error, but who is the notable exception that has more than a 10 point drop in their projection? I didn’t see it in the next section…
Bellinger is about the only primary position player I don’t see listed in the table, but there is no way his projection has dropped 10 points, seeing as his player page shows a ZiPS ROS wRC+ of 132. There is no way his preseason projection was a wRC+ of 142, right? That seems too high.
I think he was referring to Dustin May. It’s a little confusing, but the sentence can be read as referring to both the position players and the starting rotation.
Am I reading this correctly, after 96 Games and out of first by 8 games the Dodgers would still be favoured to overtake the Giants?
Implying the Dodgers to be a 94 win team this would mean the Giants are a 73/74 win team.
Feels incorrect.
Interesting. Assumptions: 96 games played. 66 games left.
After 96 games played, SF is at, say, 56-40. LAD is at 48-48.
A 74 win TT team is a .457 W% TT team
66 x .457 = 30
SF wins 86, total.
Meanwhile, a 94 win TT team is a .580 W% TT team.
94 x .580 = 38
LA wins 86, total.
At the season’s beginning, didn’t many–even most–projection systems have a 20-game (or greater) margin between SF and LA? (A 25-game margin wouldn’t have surprised me, fwiw.) Granted if you change the records that we’re assuming after 96 games played, while keeping the 8 game margin, you get slightly different outcomes. There’s also the problem of whether a team that goes something like 56-40 was in fact a 73-74 win TT team as of the start of the season (similiarly a 94 win TT team going 48-48), but basic calculus and a few reasonable assumptions should resolve that.
They did; most projection systems saw the Giants as an approximately-500 team and the Dodgers as a low-100s win team, so like roughly 20 wins.
It isn’t 94 vs 73-74, its 104 vs 84.
And that’s kind of the point of this article, but we can go a step further because we’ve seen exactly this phenomenon with the Dodgers in recent seasons. In 2018, the Dodgers started 16-26 before going 75-45 (.625) the rest of the way. In 2017, they started 12-12, had a 1-16 stretch in August/early September, and were 91-30 (.752) over the remaining games.
When the reasons for the losses do not produce lowered expectations for the players (or, as we see here, lowered expectations for some and increased expectations for others), our overall expectation of their quality also shouldn’t really change.
I want to say I respect others varying opinion of the game and the advanced stats that come with them. I even recognize the importance of why advanced stats were invented to make informed decisions on future performance and team success. Even knowing, recognizing, and acknowledging all of this, what I can’t respect, what my mind will never get around, is the absence of personal security in one’s ability to use good and effective judgment with the eye test. In this, I mean using the eye test and no being so insecure and untrusting in one’s own ability to judge objectively without the need of stats all the time.
In our society we’ve studied things such as rocket science, quantum physics, chemistry, computer science, engineering etc. All of these subjects used qualitative observations in addition to quantitative observations to make decisions.
The reason I get mad at baseball and scouts is that they act like they have no ability/trust within themselves to use any kind of effective human eye in game judgment. It’s as though we can’t ever learn anything on the fly or discover anything new or meaningful through careful watching of every situation within a game and that a stat sheet is always needed and right when it’s not. The whole purpose of our 6 senses is to gain more information all the time that changes by the day. That’s why I’m not a big fan of the big picture evaluation of players, but the short term picture since change we observe is constant and consistency is subjective. This is not me trying to imply there aren’t benefits to long term, because there are, but i find it very silly when people act like short term doesn’t or never counts because it does. Short term does count irrespective of what is believed to be considered relevant or not.
I mean if I asked someone in here to tell me what they think a consistent hitter is without consulting stats and just looking at players they observe daily, who would you say that it is?
I just feel people should be able to give me some sort of answer off of good eye judgment game in and game out as to who that would be in their mind.
I’m not necessarily expecting miracles all the time by only looking, but to think we’re incapable of judging well at all just by looking at what we are seeing is just plain silly. There is an element to human judgement and gut whether we recognize it or not, people just don’t have any faith in themselves.
But it’s not an either/or situation. Recognizing what the eye is good at seeing and what it’s BAD at seeing (either due to limitations or inherent biases) is just basic self-awareness, not cowardice. Supplementing those observations with observations that your brain can’t make on its own allows you to then compile the information effectively (“analysis”). Neither you nor I have ever heard a credible analyst say, “I can’t possibly tell you what I think” nor have they said, “Well, that looks like a home run, but let’s see what the computer says.”
Consistent hitters? Easy. Mike Trout! Freddie Freeman! Justin Turner! It’s not hard to do that and doesn’t require a PhD in statistics, but the fact that those guys are eye AND data approved should encourage, rather than discourage you…right?
It’s often difficult to know what point it is you’re trying to make, given the incoherence of your sentence structure (an objective means and standard of communication very much worth learning), but there are many reasons why statistics have become the primary vocabulary of exchange on baseball sites, including that there is little alternative. If you offer your largely subjective opinion, there’s only a very limited basis on which to reply. “X has a sweet swing. He’ll help the team for sure,” you state. “No, he doesn’t. He has a bit of a hitch that limits his usefulness,” goes the reply. Where do you go from there except to gainsay each other?
—It’s difficult to know why you think “people just don’t have any faith in themselves” when what educated people have strongly tended to conclude is that scientific observation and statistical analysis make for far more interesting and illuminating conversation than do merely personal observations and largely unsupported assertions.
In fact the best writers (on baseball as well as on less important subjects) combine the intensely personal with the intersubjective. Like scouting and statistics, it hardly needs to be just one or the other.
I think the thing that you omit here is that the theory of statistical analysis in general is that while our eyes are good at seeing things, and our brains are pretty good at interpretting them, human beings are subject to -predictable- cognitive errors in the evaluation of observations. We give too much weight to outliers and exceptions and recent events; even when we are TRYING not to let them do that, we still do so. We also can have a hard time separating the things that are predictive of the observed result from the consequences of an observed result. The Dodgers have played like dogshit; I’ve barely watched a few of these games, and I watch every game they play. But a lot of the way they look like crap are consequences (ie body language, etc) not causes of their poor outcomes. And their recent results are not as predictive of their future results as we, as humans, innately expect them to be. We over-weight what we’ve seen and what we’ve seen recently.
That’s why we have tools to help us contextualize what we’ve seen and recognize when the basic wiring of our brains is leading us astray.
I can be aware of everything I just said and I still know that when I watch the Dodgers play like that I -feel- like there’s no chance of them turning it around any time soon because they look like shit. But I intellectually know that my judgment on that issue is not accurate because my brain simply doesn’t base it off the correct predictive elements, and that a turnaround like they did in 2018, or a return to form like they made in 2017 after their 1-16 stretch that year, is the most likely outcome because they are, in fact, extremely good at baseball.
Mikolas was in the NPB, not KBO
Or, very simply, flip Garrett Cleavinger’s record as of the first week of May (0-3, 6.28 ERA, 7.95 FIP iirc), and LA is tied for best record in the majors.
When a team this good is foundering on five innings from one fringy reliever who also managed to lose two games despite giving up no earned runs in those games, and lost a third while giving up one ER, things are unlikely to be genuinely bad.
Luck happens.
And again this year they are having problems with LHP.
We’ll know a lot more by the end of May. The Giants and Dodgers tee it up seven time by then.