Why Does ZiPS Hate the Milwaukee Brewers?
As the caretaker of the ZiPS projection system, I answer a lot of questions about both how it functions and the numbers that it spits out. One question I get a lot is why the system has consistently underrated the Milwaukee Brewers, which it has over the last five seasons and by a significant margin. While I’ve talked a little bit about this issue, mostly in offhand remarks in chats and on social media, addressing that question in detail is probably necessary at this point. Of course, ZiPS isn’t alone in underrating the Brewers. But as the system’s sole developer for nearly a quarter of a century, I have a responsibility to both be as transparent as possible and improve the model as much as I can.
So, how has ZiPS done with the Brewers historically? Well it turns out that since the system was first developed, worse than it has with any other major league franchise! Here are the results for ZiPS vs. Reality since 2005. I’ll note the columns don’t precisely add up, as ZiPS projects full 162-game seasons (or a 60-game one in the case of 2020) and there are a bunch of times that teams played 161 or 163 games:
| Team | Preseason ZiPS Wins | Actual Wins | Miss |
|---|---|---|---|
| Milwaukee Brewers | 1655 | 1725 | -70 |
| Los Angeles Dodgers | 1823 | 1890 | -67 |
| New York Yankees | 1831 | 1893 | -62 |
| Houston Astros | 1631 | 1688 | -57 |
| Tampa Bay Rays | 1686 | 1717 | -31 |
| Cleveland Guardians | 1709 | 1731 | -22 |
| Texas Rangers | 1621 | 1642 | -21 |
| St. Louis Cardinals | 1764 | 1782 | -18 |
| Miami Marlins | 1486 | 1502 | -16 |
| Atlanta Braves | 1734 | 1747 | -13 |
| Philadelphia Phillies | 1699 | 1712 | -13 |
| Seattle Mariners | 1605 | 1609 | -4 |
| Toronto Blue Jays | 1676 | 1677 | -1 |
| Los Angeles Angels | 1683 | 1681 | 2 |
| Athletics | 1625 | 1623 | 2 |
| San Francisco Giants | 1665 | 1660 | 5 |
| Chicago White Sox | 1549 | 1543 | 6 |
| Boston Red Sox | 1791 | 1781 | 10 |
| Minnesota Twins | 1637 | 1624 | 13 |
| Baltimore Orioles | 1544 | 1527 | 17 |
| Detroit Tigers | 1635 | 1613 | 22 |
| Cincinnati Reds | 1593 | 1570 | 23 |
| Pittsburgh Pirates | 1511 | 1488 | 23 |
| Kansas City Royals | 1499 | 1474 | 25 |
| San Diego Padres | 1640 | 1606 | 34 |
| New York Mets | 1706 | 1671 | 35 |
| Arizona Diamondbacks | 1633 | 1592 | 41 |
| Colorado Rockies | 1529 | 1482 | 47 |
| Washington Nationals | 1624 | 1576 | 48 |
| Chicago Cubs | 1714 | 1664 | 50 |
One source of error that’s really difficult to control for is what a team does at the trade deadline. Many of the teams that have overperformed their preseason projections have added talent during the season; conversely, underperformers have a tendency to trade talent away. That’s challenging to model, since it involves trying to project players who aren’t currently in the organization as part of the team, even though we have little idea who those players will actually be four months in advance. I actually created a model based on team quality, age, payroll, recent record, and trade history to get an idea of the likelihood a team will be a buyer or seller in an upcoming season. But while it sort of works, its accuracy isn’t up to the level where I’d include it as part of a projection.
Historically, the Dodgers and Yankees have been two of the league’s most aggressive buyers, so it isn’t surprising to see them atop the list of the biggest ZiPS misses. But while the Brewers have made some big in-season moves — the biggest arguably being the CC Sabathia trade in 2008, which was one of the most effective trades of this type ever — they aren’t on the buy side as frequently as some of the other underprojected teams. So, what’s going on here?
First, here’s an overview of how the percentiles for team projections have worked out. Ideally, you want 10% of teams to exceed their 90th-percentile projection, 20% of teams to exceed their 80th, and so on:
| Percentile | Percentage of Teams That Exceeded |
|---|---|
| 90th | 9.3% |
| 80th | 21.0% |
| 70th | 29.8% |
| 60th | 41.5% |
| 50th | 50.5% |
| 40th | 58.8% |
| 30th | 69.1% |
| 20th | 78.4% |
| 10th | 88.9% |
ZiPS does a pretty good job in the aggregate. To put it simply, the basic job of a projection system is to know the range of possible outcomes, and be wrong by the appropriate margins the proper number of times. It would be easy to say “Hey, projections work as they’re supposed to in the aggregate, and some team is inevitably going to have the worst projections of the 30, so whatever,” but that doesn’t mean that we shouldn’t investigate these issues and assess whether there’s something systemic that the model is missing. Especially so in a case like Milwaukee, where nearly two-thirds of the 21-year error comes from the last five seasons (417 projected wins vs. 463 actual wins).
The ZiPS projected standings have two components: the projections themselves and the estimates of who actually ends up with playing time. To get an idea of how much of the ZiPS misses are errors in projection compared to errors in playing time, I will frequently re-project team wins using the actual playing time for each player after the season is done. Re-projecting the 2021-2025 Brewers using their preseason projections but the players’ actual playing time makes the issue a lot clearer:
| Year | ZiPS Preseason | ZiPS Knowing Actual Playing Time | Actual Wins |
|---|---|---|---|
| 2021 | 83 | 93 | 95 |
| 2022 | 88 | 94 | 86 |
| 2023 | 84 | 87 | 92 |
| 2024 | 78 | 87 | 93 |
| 2025 | 84 | 90 | 97 |
| Total | 417 | 451 | 463 |
Knowing each player’s actual playing time doesn’t eliminate the errors, but it whittles the missing 46 wins all the way down to 12. In other words, ZiPS isn’t doing a bad job with the projections; Dan Szymborski has done a poor job guessing which players will end up with playing time for the Brewers! Injuries are sometimes a reason for playing time discrepancies, but they typically result in teams underperforming their projections as regulars miss time. Not only have the Brewers overperformed, they’ve done so while not being particularly good at avoiding injuries; they’ve actually lost slightly more wins than the average team due to IL stints over the last five years.
Instead, what appears to be happening is that the Brewers have been extraordinarily successful at giving more playing time to players exceeding their projections. For example, there were 62 hitters who had seasons with at least 200 plate appearances for the Brewers from 2021 to 2025. As a group, ZiPS only underestimated them by 1.5 points of wRC+ in the aggregate (104.7 actual vs. 103.2 projected). But of the 33 hitters who exceeded their projected wRC+, 28 of them received more plate appearances than I had as my baseline expectation. The same is true for pitchers, especially relievers. Now, there’s a natural tendency for teams to give more playing time to players who are outperforming their projections and less to guys who are underperforming, but the Brewers have been notably more successful at this than the rest of the league. From 2021 to 2025, 81% of their qualifying players who outperformed their expected wRC+ or ERA+ got more playing time than I expected as a baseline. To put that into context, the league-wide rate was just under 61%, and no other team was above 70%.
So, how do I fix the Brewers’ projections? That’s a bit of a craggy problem that I’m still working on. This offseason, I tried to be more aggressive in my assumptions about who would get playing time for Milwaukee based on the quality of their projections. As a result, ZiPS forecast the team for 85 wins. Naturally, the Brewers are on pace for 99.7 wins as of Wednesday morning. I may need to more accurately project actual front offices; if the Brewers are simply better than everyone else at evaluating their talent with information only they have access to, it’s not something I can directly correct for. Unless, of course, the Brewers decide to just give me all their internal data, which seems unlikely. Or if I, say, catch Dan Turkenkopf in a giant net and imprison him in my tool shed until he spills the beans. As much as I like improving projections, I don’t think my employer would appreciate if I did so by committing federal crimes, so I’ll simply have to keep trying. Being wrong is how we improve predictive models, and let’s just say that the Milwaukee Brewers continue to give me a lot of opportunities to learn.
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.
Does this mean Cubs are terrible at giving playing time to those who outperform their projection?
Don’t sell the Cubs short; they could be terrible at lots of things!
I was gonna say, this buried the lede on ZiPS being an irrational Cub fan.
This is fantastic info, thank you
Yeah this type of visibility into the causes of error for a private model is unparalleled. I literally cannot think of a similar example (but would love to hear about other places I can find this sort of work on private models)!
Reading this type of work makes me a much better thinker and modeler, and teaches me so much about the craft. I’m so grateful to Dan (and Fangraphs, and the whole team) for teaching me so much over the years I’ve been reading.
Yet another reason why Fangraphs is the best baseball website in the business, and one of the best websites out there, full stop.
The current successful run has occurred after the 2018 renovations to the Brewers’ Am Fam Fields of Maryvale. The locker rooms, agility fields, gym, etc were all significantly improved for the players. Also, they doubled the office space. It’s likely this investment in organization has allowed Dan Turkenkopf & company to find those additional wins.
What year did they hire August Fagerstrom?
The 2016 / 2017 offseason.
This is not actually a bad theory. August Fagerstrom is an incredibly talented guy, and is likely familiar enough with models like ZiPS that he could create his own.
How long have I been on fangraphs?!?!
A Carson Cistulli/Jeff Sullivan amount of years?
Yes, Dan, it IS the Brewers front office!
Dan, two things:
1)”I don’t think my employer would appreciate if I did so by committing federal crimes,” What even is a federal crime these days anyway?
2)I’m sure this is a dumb question and the answer is yes. But, have you tried doing a retroactive study where you try to repredict some portion of the past 5 seasons? Specifically, I wonder if the Brewers are just better at applying bat tracking data that wasn’t publicly available until the 2024 season. How would what we know now have altered your last 3 years of predictions?
For 2), I’m fairly certainly that’s how Dan tests new inputs into his model, however I haven’t seen him publish any “retroactive” projections beyond one off ones for players for specific articles
Same, that’s why I started off by saying it’s probably a dumb question.
Why does ZiPS hate the Dodgers and the Yankees?
More importantly, why does it love Toronto, the Angels, and Athletics?
I think this says that it’s most accurate on those three teams – not that it has an irrational crush on them (like it does for the Cubs)
Dan mentions that they are frequent deadline buyers, but there’s another factor. Computer models don’t usually predict extreme win totals. The Dodgers have 5 100-win seasons in the last 10, and their 2020 record was even better on a percent basis. I doubt Zips had the guts to call 111 wins!
Is aggregate wins over 21 years the right way to do this? Sure, the article is about the Brewers out performing for the last five years, but wouldn’t a better way be to take the absolute in difference for each year and add them up? Maybe ZiPS is really crappy projecting the Blue Jays but it just swings back and forth each year between over and under projecting wins.
I’m not so sure. Suppose I predict 87 wins for the Blue Jays each year, and they run totals of 80 94 80 94 80 94. ASSUMING they are stable on talent year-to-year, then they are an 87-win team on talent, just random variation each year is drifting them above and under that center expectation. There will probably always be some of that unpredictable random variation, so I crushed the nail on their head in what I could control, capturing their actual quality across the stretch (while acknowledging the potential for variation).
Either way, I’d be curious what your approach’s results would be versus the presented.
I think the percentile data more or less cover this. If the Jays were 10th percentile one year and 90th the next and it wasn’t random variance just happening to hit a couple of times in the same place, we could expect to see those tails rise up well above ~10% incidence.
Remember, I’m not really talking about *accuracy* here, but *bias*.
Running the table from 2005 to 2025 is burying the lede somewhat since the Brewers projection beating really took off in 2016 when Stearns/Arnold arrived.
I know that ZiPS is only half of Depth Charts (maybe more like a third…ZiPS, Steamer, playing time allocations) but the Brewers are +87 Wins over their preseason win projections on the Playoff Odds page for the nine completed full seasons from 2016 to 2025.
Eight of nine years over, with five seasons of +12 or greater, and on pace for another double digit overage this year.
We know the Brewers have been gaming run prevention for a decade now…
2016-22: (93 ERA- | 6th) and (-0.16 ERA/FIP | 4th)
2023-present (86 ERA- | 1st) and (-0.41 ERA/FIP | 1st)
That’s an MLB best 3.91 RA/G over their last 538 games (3.46 so far this year) and Depth Charts projects 4.25 RA/G (8th lowest) rest of season.
More interesting though might be the progress their offense has been making over the last two tree years since Murphy took over and they started developing position players too (an MLB best 52.6 WAR from players 28 and younger since 2024 with NYY in second at 41.2 WAR including their eight win Soto boost)…
2018-2023: (3,961 R | 14th) and (100 wRC+ | 16th)
2024-present: (1,940 R | 4th) and (105 wRC+ | 9th)
Calendar Year: (831 R | 1st) and (111 wRC+ | 4th)
So 4.89 R/G over their last 376 games (4.94 so far this year) and Depth Charts projects 4.43 R/G (14th) rest of season.
I’d expect some regression from their current levels no doubt, but they haven’t allowed as many as the 4.25 R/G or scored as few as the 4.43 R/G Depth Charts is currently projecting since 2022.
I think it’s likely based on the description here and elsewhere that the Brewers and other over-performers have models that are similar to ZiPS and that mean they play their better players more often.
But the idea of modeling playing time separately by team based on the projections themselves sounds like it would require a nightmarish set of assumptions.
I don’t see any way to do this without overfitting and screwing up the projections more for other teams.
I wonder if they have internal thresholds established for their highest variance players: “if player x hits for a wOBA of y over z PA, he gets more playing time” or more likely “if player a has a swinging strike rate below b over c PA, he gets more playing time”
Something like pairing that type of sensitivity analysis with better communication between the FO and the dugout would help to sort out what is hot hand noise and what is true talent improvement that should be used to reset playing time allocations going forward. Operationally, that seems like the best way to beat projections in the way they have.
Great article
And thanks for the transparency
“Why does Dan hate my [favorite player/team]: The Series” is maybe too much fan-service, but if there’s one man suited to do it…
“What appears to be happening is that the Brewers have been extraordinarily successful at giving more playing time to players exceeding their projections.”
Might this imply that the Brewers’ internal projections are some of the more accurate in the league?
Or perhaps they’re less married than most to the old sporting tropes of:
“You shouldn’t lose your job because of injury”
Or, “He just needs to work through it and will come around”
etc etc
So it might be less that they have better projections, but more they believe them more and are willing to act on that belief.
If so it hasn’t transferred to the Cubs and Mets, or maybe they gave Counsell and Stearns wrong information.
A whole org has to buy into a philosophy like that, top to bottom. You can’t just hire one person and expect every other employee to change how they think about their work.
Neither the cubs nor the Mets actually wanted to change how they worked – they just wanted the benefits of better work and thought they could buy it without changing themselves.
Classic rich guy shit.
I think I just prefer the easiest explanation of, Players Who Go To The Brewers Just Get Better.
Yeah I mean why think about stuff right?
I have. And the explanation I’ve come up with is the Brewers are better at developing players than everyone else. Dave’s tried to come up with an explanation in chats and now in this article, but the simplest explanation is probably the most accurate: players go to the Brewers and get better. That’s going to break statistical models until models start accounting for it.
Thanks for explaining that.
Or until the league catches up, personnel move on, etc. I don’t see how a model would account for the strength of an organization’s development quantitatively.
Well, you could look at the delta of a player’s projections year over year. Do players who go to the Brewers actually improve more than league average? Are there particular player types or skills that improve more than league average? Or are the Brewers just better at making earlier calls about which players to put on the field as skills change at rates comparable to the league average?
It’s there enough data to draw conclusions? How many players per year go to the Brewers from somewhere else? And how do you account for the quality of development they’re coming from? Eg. Vaughn? White Sox dev sucked. Credit to MIL or debit from CHW?
and then, step 2: start correlating those improving players with the coaches, scouts, etc that are involved with said improvements
I’d be curious about how much the Brewers spend on things that effect player development compared to other teams. It seems like such an easy area to invest a few million dollars to get a significant advantage over the half of the league that’s more concerned about saving money anywhere they can, while also saving you tens of millions in the process.
We’re all curious. It’s not public data though.
But doesn’t he address that in the article. About 3/4 of the difference is driven by playing time differences, not players outperforming their projections.
As someone who has watched at least 50 of Tyler O’Neill’s 26 strikeouts this season en route to a .158 batting average in his first 107 plate appearances, this is a fascinating source of edge for the Brewers.
Yeah, well the Brewers also have given 163 PAs to Luis Rengifo, who is almost certainly one of the 20 worst position players to get any real playing time this season. Nobody is perfect!
Now I’m wondering what their model likes about him
I don’t think it necessarily likes him. The Brewers have done this with veteran infielders before. They’ve got some infield talent ready to come up from AAA soon (Cooper Pratt and Jett Williams) so Rengifo is keeping a spot warm for them. If he played well then they’d figure something out but if he doesn’t then they start moving him out when the AAA guys are deemed to be ready, which should be soon.
Also the kind of guy they’ve often replaced midseason. The article calls the Brewers non-buyers, but that’s not exactly true. They’re not picking up big name players, but they almost always make a couple moves at the deadline.
It doesn’t have to like anything about him per se. Merely that he’s not worse than Caleb Durbin, and adding Kyle Harrison was worth the switch.
Rengifo has a career wOBA of .297 and career xwOBA of .303. He has a .243 wOBA this season with an xwOBA of .296. He has been one of the unluckiest position players to get real playing time. Every time he does hit the ball hard it goes right at someone.
He isn’t a good hitter in general and I can’t wait for Jett to come up but he has not been some huge problem from a process standpoint, just a results one and in these tiny samples the results have a ton of luck in them.
Thanks Dan. As a huge Brewers fan who follows the team closely, I would posit that what causes the Brewers (lately) to outperform projections is there depth and relative lack of “stars”. Take today — Logan Henderson couldn’t pitch because of back tightness, so Chad Patrick and Drohan covered the starter’s innings by giving up 1 run. They are excellent pitchers and both would probably be in the starting rotations of the majority of teams. And now they are calling up Coleman Crow for a “spot start”. Coleman Crow is a pretty good pitcher who also would be starting for a lot of teams but is probably 8th on the Brewers depth chart. And the Crew certainly missed Andrew Vaughn when he was out, but the drop-off from him to Sanchez/Bauers just isn’t that great (nor the dropoff from Yelich to Bauers). This isn’t just playing the right guys. This is having guys in reserve that can play. So I would suggest that, knowing that injuries are bound to happen but are unpredictable, you should add a component to ZIPS (assuming you haven’t), that takes organizational depth into account. Maybe run an analysis of what would happen to a team if they lost 4 of its best players for a month or so and see which teams would have the greatest and least drop offs for expected wins.
It is definitely easier to make good choices when your choices are mostly good.
I get your broader point about depth, but this description doesn’t match the source of projection error described in the article. If all the guys in reserve “can play,” then it shouldn’t have mattered so much who Dan projected to receive the playing time vs. who actually got the playing time. It seems clear from the article that the Brewers are doing something that makes them better at “internal scouting” or whatever you want to call it than ZIPS, or than other teams — not just accumulating good players in larger quantity (though that’s obviously part of it) but also knowing them well enough to play the right ones more.
I’m pretty sure ZIPS does that already. Some number of the thousands of runs include injuries to Players A, B, or C and goose the playing time of their backups accordingly.
And I could be wrong, but I think this year Dan said he made a concerted effort to put more weight toward the Brewers’ depth options because they had established a high likelihood pattern of using them.
Hesitate to sound like “Intangibles Guy” on a Fangraphs page, but Contreras and Co.’s ability to call savvy sequencing/location, manage pitcher emotions (OK, except Uribe, lol), trust the plus defense enough to throw strikes in tough situations, etc. really seems to play a possibly unmeasurable but meaningful role in wringing the most out of whichever 13 pitchers are rostered at a given time. Idk, but I do suspect that a fair amount of run prevention is not currently being captured by stats, and that we would be wise to avoid assuming a framing metric adequately reflects a catcher’s true defensive acumen just because it’s all we have at our disposal.
I have the feeling people will talk about Contreras the way they talked about Molina. A big “intangible” in terms of game calling, etc…
“relative lack of ‘stars'”
Nobody else seems to have mentioned this, but there is something going on here. Call it the “Machado” effect.
I’m picking on Machado because like Tyler O’Neill mentioned above, he is having a horrible season but is still out there hacking away. Every. single. day.
What are the Padres supposed to do with a guy making $32M/yr. who is barely performing at replacement level? He is not even hitting well enough to pinch hit.
Among teams that regularly win games, this situation seems to be least likely to occur on the Brewers. Who are their Machados and O’Neills? There aren’t any players on the roster who they would trot out there every day packing a wRC+ of 73.
That has got to help exceed expectations over the long haul.
I compared the Brewers to the Carolina Hurricanes prior to the season (at the time, largely because they are not married to size and traditional ideas of what players *should* be – i.e. Contreras was a bad defender at C? Are you sure about that? Why cant we just make him a good one? Why cant this CF prospect just be an excellent defending COF, etc) and I think this is another way they are similar
Due to how they both run their organizations, they turn over their rosters adhering to aging curves and thus are not ever “pot committed” to poor performers making a lot of money
This is of course a double edged sword, as every postseason failure gets blamed on a lack of Stars performing when the lights are brightest but it guarantees you will be back in the postseason the next year minus one or two vets replaced by a new group of younger guys. The only player on their roster who they would have to play if he was having a rough go of it is Yelich, and he’s managed to stave off decline well enough to be a contributor
edit: Interestingly, this probably means if the MLB Owners win the power struggle, it’s going to remove some of the Brewers’ competitive advantage since they’ll have to pay some of their Freddy Peraltas and Willy Adames’ to stay
This is my ignorance so I’m asking with my hands up, but has baseball analytics come around on base stealing yet? Like most of you I read Moneyball 20 year ago and I remember stealing bases being a big no-no. Forgive my ignorance, but I don’t remember reading about a shift back in the decades to follow. But right now the Brewers, Guardians, and Rays are in the top five in SB and they all lead their divisions. Marlins and Nats make up the other two and they’re performing relatively well.
So honest question, have projections come around on base running aggressiveness?
It was never that stealing bases was a no-no, more that you had to be way better at stealing than most people thought at the time for it to outweigh the cons. It shows up in WAR when guys are especially good at baserunning.
The lower run environment has made the base stealing success rate lower in recent seasons. These teams may be a bit ahead of the curve on that, or (more likely) the types of players who steal a lot of bases are also contributing in other ways that adds up as they are likely elite athletes
If you want to know how fangraphs values base running, go to team stats and sort by BsR.
One thing you should know is that it’s evaluating base running with a lot more than who has the most stolen bases. It’s taking extra bases, it’s not wasting outs on the base paths, etc.
But either way, it’s not a lot. The best team is about ten runs better than the worst for the season. Smaller spread than you see in defense or hitting.
Re: the difficulty of projecting which teams will buy or sell at the deadline, what does the ZiPS error distribution look like if you only look at projections up until the trade deadline?
I have a similar question. If I remember correctly, ZiPS is updated daily, so could you determine error numbers by using the preseason projection up until the deadline and then post-deadline projections for the remainder of the season? Of course the post-deadline projections would have a lot more info that the preseason ones, but it might still yield some insight into which teams are over/underperforming relative to the talent on the team (in the eyes of the model).
Would you opine that the Brewers do in fact have private information allowing them to outperform the projections or are they just the largest benefactor of normal variance over the past twenty years? You’d think Stearns would’ve brought that magic wand to the Mets.
And Counsell to the Cubs
Have to assume this advantage comes from the kind of organizational strength that doesn’t inhere in one person. A strong scouting/development staff with a well-worked-out internal structure and culture, or a proprietary data/modeling thing, or maybe even both. Institutional knowledge and practices that take a lot of time and turnover to change.
Here’s a question that might be worth some follow-up: do the Brewers have more overperformers that the model would expect, or than other organizations?
The explanation here makes sense — the Brewers are better than most teams at identifying their best options and giving them playing time. Which, incidentally, might be one area where it helps that they’re a small market team with more lower-salaried players — they are less likely to run into sunk cost fallacy situations and stick with higher-paid players longer than other teams might, and more likely to have roles available for overperforming players.
But it makes me wonder whether they are also better at *generating* overperformance, through identifying them through scouting or through player development or likely both. Certainly seems to be the case on the pitching side especially, but I’d be curious to know whether comparing their performances against projection benchmarks bears this out as compared to other organizations.
In other words, this analysis shows that they’re taking advantage of the opportunities overperformance creates. My question is, are they also generating more of those opportunities than we would expect?
It makes a lot of sense that so much attention is paid to ZIPS misses on the Brewers bc not only are they systemically underprojected, the Cubs (their primary division rival for most of the last decade) has been systemically overprojected, so the difference is more notable in final standings.
Meanwhile the similar projection miss w the Dodgers is effectively meaningless because they’re rarely ever in jeopardy of actually losing the division.
My memory is fuzzy, but have the Dodgers benefitted also from other divisional teams being sellers, perhaps?
So the secret sauce is to play your players who are playing well more often?
Well that seems kind of obvious!
In the same way that the secret to the stock market is to buy the ones that are going to go up, yes.
Agree with Roger. The secret sauce would be to play more often those players who are going to play well, not necessarily those who are playing well at the moment. That seems like a pretty big distinction.
Good job Dan. It’s tricky because being “more aggressive in my assumptions about who would get playing time” will either nail it or make the projections more aggressively wrong. It’s still dependent on who you pick.
As someone who watches a lot of Brewers games, a lot of what they do seems like a miracle. Earlier in the season with Chourio, Yelich and Vaughn out, the lineup did not look like a major league club. (They had a game with Lockridge, Contreras, Rengifo, Sanchez, Ortiz, Matos, Frelick, Perkins and Hamilton…that’s right, Rengifo hitting third…against Garrett Crochet).
I think Pat Murphy does two things that make playing time hard to project (and there is a third Brewers’-wide factor). First, he definitely rides a hot hand. Because he does not have a lineup full of stars, he has a lot of chess pieces to play with. To use a fantasy analogy, he has five streamers in the hitting lineup, and a few more in the rotation. Almost none of them are good enough to start every day, but they are worth a “pick up and start” when they are going good. He has no other choice, which must be fun for him in a twisted sort of way. But super hard to predict.
Second, he’s a master at maximizing what a player is good at. For example, David Hamilton is a terrible hitter but is a blazer…his HP-to-1st is 4.01 on average, which means we regularly see him under 4. So Murph has him bunting a LOT. I think he has 9-10 bunt base hits and his BA on bunts is sky high. ZIPS may be able to model his 1b and Avg from the historical data (subbing in a few bunt singles for regular singles), but how can you account for Murph believing he should be in the starting lineup because he is an invincible bunter that messes with the opponents’ heads? There is nothing in his numbers that suggests he is going to get 400 PAs or more. It’s a mind game.
The teamwide factor (alluded to in the first Murphy point) is that the Brewers have a lot of bad hitters, which means they cannot rely on what we would consider good hitting. Rengifo and Joey Ortiz are appalling hitters. (At least Ortiz is a good defender). But they are high contact guys, and what the Brewers like are runners and balls in play, and then hope for the best. That’s a forced philosophy because they don’t have the payroll to approach it any other way. They are going to be on the BB% and BABIP leaderboard in this configuration. I have not studied it, but I wonder if that philosophy messes with the regression to the mean in some stat categories. Of course, that would show up as stat projection misses rather than playing time misses, but it might inform playing time projections. If Brewers hitters are more likely to increase their BB%, or hold on to what seems like an artificially high BABIP from the prior season, it may not make a huge difference in the ZIPs projections, but might be enough that when you see the ZIPs projection, you are more inclined to allocate PAs to that player.
Anyway, it’s fun thinking about. Thanks for all your work.
Pat Murphy has only been around for a few years. If it’s a manager thing, why can’t Counsell replicate it with the Cubs?
Great post Chapel. Because the Rengifo/Hamilton/Lockridge types play solid defense and are at least capable of small balling when the situation calls for it, the Brewers can reasonably expect some positive and often timely contribution from them throughout the course of a game. That’s not saying much, of course, but collectively it’s a stark difference from Milwaukee’s rosters built in the waning years of Stearns/Counsell.
People forget that as recently as 2022, the Brewers were a station-to-station offense “led” by Rowdy Tellez, Willy Adames, Hunter Renfroe, Andrew McCutchen (old version), Luis Urias, and Caratini/Narvaez. Yelich and Wong (ex-porn stars) were the only speedy contact hitters getting much PT. [Fun fact: In 2021, Lo Cain led the Brewers in SB … with 13.]
Those teams eked into the playoffs, then stopped hitting homers vs top pitching and got unceremoniously bounced.
Murphy’s teams cruise into the playoffs but also have and will likely continue to struggle to slap and bunt their way to runs vs top pitching/defenses. The best solution is to purchase Ohtani, Freeman and Tucker. Sans that option, my opinion is the all-around fundamentally sound Punch n Judys are a more fun way to score <3 runs per playoff game than the one-dimensional lumbering whiffers.
To exemplify this rather recent shift in roster-building approach, consider …
Josh Bell checked a lot of boxes for the Brewers this past offseason. Team desperately short of power? Bell hits 20 dingers every year. 1B/DH platoon partner for both Vaughn and Yeli? Bell switch-hits. Eminently affordable FA on a 1-year deal? Check again.
If this same scenario occurred in 2023, I’d bet dollars to donuts Milwaukee lands Bell (see: Hoskins, Rhys). Instead, back in mid- November, the current front office committed the reserve “power bat” spot to Jake Bauers. Unlike Bell, he plays a plus 1B and a passable LF. He runs the bases fine. He’s 3 years younger and less at risk of a plummet in bat speed. He strikes out plenty but takes walks and extends at-bats, improving in both categories in 2025. Surely the Brewers didn’t expect Bauers to be leading the team in HR on June 1, and yet, you don’t have to squint too hard to understand why Matt Arnold could believe a well-rounded (and cheaper) Bauers would contribute more overall value in 2026 despite the 9-HR gap in their preseason Steamer projections.
Until we start quantifying (or until Fangraphs starts showing us) the data for front office talent, this will always be an issue. I’m assuming that, like the 1991 Gulf War display of the “new” Stealth Fighter (F-117) that was actually developed twenty years earlier, that quantifying front offices is the frontier of projection systems right now and we the peasants won’t see the data for some time.
Quantifying only projected player data in a vacuum will only get you so far.
I don’t think we’ll ever see it. Or not in a way that’s useful. Say they’re really good at scouting? When does that show up in the standings? 3 years from now? By then maybe half your scouts left.