Earlier this offseason, Dan Szymborski wrote about the Giants’ remade rotation and the strategy behind it. In short: Alex Wood and Anthony DeSclafani were interesting free agents, and without much chance of reaching the playoffs, the team opted to buy low and hope to catch lightning in a bottle. In the weeks since, San Francisco has repeated the same project, only this time for the bullpen, highlighted by Jake McGee’s two-year, $7 million dollar deal announced on Tuesday.

McGee joins a sampler platter of relievers who might be good on cheap deals. In early December, the Giants signed Matt Wisler to a one-year deal. Later that month, they signed John Brebbia, who will spend at least the first half of the year rehabbing from Tommy John surgery but could be a late-season bullpen piece. They’ve also extended minor league deals to Dominic Leone, James Sherfy, Silvino Bracho, and Zack Littell. Spaghetti, meet wall.

McGee is the most interesting member of this group, which explains why he got two years and real money while everyone else got only a single year. He combines three things that teams look out for in relievers: he strikes out more batters than average; he walks fewer batters than average; and he’s left-handed. Those three traits alone are enough to get a job, and indeed, the Dodgers signed McGee to a one-year deal before 2020 when he was coming off back-to-back -0.3 WAR seasons.
Read the rest of this entry »

Earlier this week, as is tradition, FanGraphs founder David Appelman went into his garage, turned off all the lights except for some candles, and performed a dark and arcane ritual. Then he went back inside, pushed a few buttons on his computer, and now we have playoff odds for 2021!

Okay, fine, that isn’t exactly how it went down, but it’s close. Our playoff odds incorporate little pieces of a lot of features you’ve already seen on the website. We start with a blended projection that incorporates ZiPS and Steamer’s rate statistic projections. We add in playing time projections from RosterResource, which incorporate health, skill, and team situation to create a unified guess for how each team will distribute their plate appearances and innings pitched.

With playing time in hand, we use BaseRuns to estimate how many runs each team will score and allow per game based on our earlier blended rate statistic projections. That gives us a schedule-neutral win percentage for each team, because you can turn runs scored and runs allowed into a projection via the Pythagorean approximation. From there, we simulate the season 10,000 times, with an odds ratio and a random number generator determining the outcome of each matchup. Voila! Our playoff odds.

Why am I telling you all of this? First, so you can look at them. They’re accessible from the main page, but you can also click here to dive in. Second, because I’m going to walk through some projections I found interesting, as well as a few places where the gap between common perception and our odds merit an explanation. Let’s get started! Read the rest of this entry »

Read the rest of this entry »

It feels like a long time ago that the Oakland A’s won the AL West. Partially, that’s because everything feels like it lasted forever in 2020 — the last year has been the longest decade of our lives. Partially too, though, it’s because the team has spent the vast majority of the offseason doing nothing.

Marcus Semien, Liam Hendriks, Joakim Soria, Yusmeiro Petit, Mike Minor, and Tommy La Stella all left in free agency. On Saturday, the A’s made their first major transaction of the offseason, trading Khris Davis and Jonah Heim for Elvis Andrus. Later that day, they issued their first (!) major league contract of the offseason, signing Mike Fiers to a one-year, $3.5 million deal, as Ken Rosenthal and Alex Coffey first reported.

Fiers was an important but unsung part of a run-prevention monster in 2020. He made 11 starts for the A’s, who reached the playoffs on the back of a simple strategy: prevent some runs with starting pitching, hold some leads with a phenomenal bullpen, and sprinkle excellent defense around it to make it all play up. Fiers wasn’t an ace, and he didn’t need to be.

That appears to be Oakland’s plan again this year. Their projected rotation is close to unchanged (Minor started a handful of games, but the other five top starters are all back):

Fiers will be the worst everyday starter, like he was last year, but he’ll take the ball every five days, and the A’s would have struggled on that front otherwise. Daulton Jefferies, A.J. Puk, and Grant Holmes are the next three starters up, and all have had serious injury issues in the past few years; the team will likely want to limit all of their innings this year. Read the rest of this entry »

The top free agent pitcher in baseball is no longer a free agent. After an interminable PR tour, Trevor Bauer has signed with the Los Angeles Dodgers, as Jon Heyman first reported. The deal, a three-year, $102 million pact with opt outs after each year, bolsters an already-stout Dodgers rotation and ups the NL West arms race after the Padres’ busy offseason.

It would hardly be honest to write about Bauer without mentioning who he is as a person, so let’s do that first. For lack of a better way to say it, he’s a jerk, a troll. That’s not harsh enough, but it points in the right direction. “Troll” undersells it: time and again, Bauer has stepped up to the line and then gone past it, lashing out and inducing his fans to harass someone before acting shocked at the fallout, claiming innocence.

I won’t detail each individual incident, but suffice it to say that this goes beyond your typical Twitter sniping. The pattern is shockingly similar each time: Bauer takes offense at some perceived slight on social media, berates and otherwise insults the source of that slight (sometimes at great length), and then with a quote tweet, points his fans and followers in the woman’s direction (and it’s almost always a woman), who then proceed to harass her.

Eventually, Bauer issues a banal non-apology about how he never intended to harm anyone and doesn’t believe he did anything wrong, despite the glib falsity of that statement. This isn’t an isolated incident, a poor decision made in his rash youth. It’s a pattern, and a well-documented one.

I’m not here to legislate how you feel about that. I’ll simply invite you to consider how it feels to root for someone who repeatedly takes advantage of his popularity and power to make life worse for people without those things; how it feels to be one of those people. For the remainder of this article, though, I’m going to talk about what this means on the field, on the days where Bauer is pitching, though that hasn’t always been without conflict either. Read the rest of this entry »

I’m going to spend about half this article trying to deceive you. I just thought you should know that upfront. Why? Alex Chamberlain wrote a wonderful article about Kyle Hendricks yesterday, and it reminded me of one of my favorite paradoxical findings about pitching. Now, I’m going to use that finding to bamboozle you — or at least, that’s the plan.

Hendricks, you see, is spectacular at throwing pitches in the shadow zone, the boundaries of the strike zone and the area just outside of it. That’s an obviously useful skill. When you watch a pitcher painting the corners, it doesn’t just feel like hitters are unlikely to make solid contact, it’s actually true. Batters have worse outcomes on the borderline part of the plate than in any other zone.

Here’s a statement that I don’t think is at all controversial: pitchers exercise a lot of control over where they throw the ball. It’s not like exit velocity — mostly batter-controlled — or walk rate, which depends on myriad factors that both pitcher and batter affect. Where a pitcher throws the ball should be up to, well, the pitcher.

It stands to reason that if pitchers control where the ball goes, there will be leaders and laggards at it. As it turns out, there are! As Chamberlain pointed out, no one has thrown a higher percentage of their pitches in the shadow zone (since 2015) than Kyle Hendricks; Marco Gonzales is narrowly behind in second place. Sure sounds like a skill to me. Read the rest of this entry »

This week, one NL Central team has acquired the best defender at a key infield position. He’ll be playing the 2021 season at age-30, and he was below average offensively in 2020, so it’s not as though there aren’t red flags, but great defense doesn’t grow on trees. That’s right: the Brewers signed Kolten Wong to a two-year, $18 million contract, as Jon Morosi first reported.

While Nolan Arenado might have fallen behind Matt Chapman in the third base defense hierarchy, Wong reigns supreme at second. For three straight years, he’s won the Fielding Bible award at second base. Every advanced defensive metric sees him as the best fielder in the game over the past three years. Bigger fan of the eye test? He can do this:

Okay, fine, single defensive highlights are a bad reflection of talent. But he can do this, too:

I could go on all day if there weren’t an article to write. Wong is one of my very favorite players to watch. This is a transaction analysis, though, so I’ll restrain myself, and merely say that our very good defensive projections for Wong might still be conservative. Read the rest of this entry »

Hey everyone, and welcome to the convergence of two recurring segments. It’s the highly awaited crossover between “Can You Believe the Cardinals Got Nolan Arenado for That?” and “Let’s Value Gimmicky Contracts,” two columns I almost assuredly enjoy writing more than you enjoy reading.

Let’s get the deferred money part out of the way first, because while it’s obviously very important to the Rockies and Cardinals, it has nothing to do with Arenado’s decision-making. He’s getting his cash, and whether the check says Monfort or DeWitt, the cash still spends the same. It won’t affect his decision on whether to rip the whole contract up.

As Jeff Jones reported, Arenado agreed to modify his contract as part of the trade. In 2021, he was due $35 million. Now, he’ll receive $15 million this year, paid directly by the Rockies. He’ll also receive $20 million in deferred compensation, regardless of whether or not he opts out. If he’s still under this contract, that money will be sent to the Cardinals, who will then pay it to Arenado. If he opts out, the Rockies will pay him the $20 million directly.

Finally, if Arenado doesn’t opt out, the Rockies will be on the hook for the $16 million he’s due in 2027. That’s the new year that the Cardinals agreed to as part of the trade, and while it’s unclear exactly why the Rockies chose to pay that part rather than some pro-rated portion of earlier salaries, here we are.

For the Cardinals, this is a great fit. They’d been acting as though cash was a key constraint this year, and getting a year of Arenado at no cost (literally, no monetary cost!) does a good job of making the short-term books work. In the long run, they’re paying him nothing for one year (if he opts out after 2021), $35 million over two years (if he opts out after 2022), or $164 million over seven years.

With that covered, let’s talk about Arenado’s options. The structure is as straightforward as it gets for these kinds of things. After 2021, Arenado will have the option to walk away from the entire deal and become a free agent. If he opts to remain in St. Louis, he’ll get another chance to wash his hands of the deal after 2022. If he still wants to stay, then he’ll be under contract until after the 2027 season.

When Arenado signed his extension two years ago, I covered a probabilistic way of thinking about the value. In the interim, a few things have happened. First, Arenado had a down 2020. Second, Dan “Dr. ZiPS” Szymborski gave me a long-term forecast for Arenado that beats my generic aging expectations. Finally, I’ve added a few bells and whistles to the option model in the interim. For the most part, though, we’re just running it back.

To calculate the value of an opt out, we need a few things. First, a central projection for how good a player will be in the future. ZiPS is all over that. Here’s the next five years of Arenado’s projections:

I’ll assume a standard aging curve after that, which gets us a central tendency for how his career will go.

Next, we need to apply variance. I’ve found in previous studies that projections for players who fit Arenado’s mold — mid-career and projections above 2 WAR — vary with a standard deviation of roughly 1.4 WAR from year to year. We’ll start with projections as a baseline, but to figure out the value of an option — the right but not obligation to do something — we’ll have to simulate 10 million or so different futures, then work out what happens on average.

Finally, we need to figure out how to translate Arenado’s projections into a potential new contract. Converting WAR to dollars misses some team-building effects, and as Craig Edwards showed last year, it’s not as simple as applying a linear conversion. Still, it’s my model, and it’s just for Arenado, not for the dang league as a whole. Let’s start with $8 million per win.

While we’re varying projections, we also need to vary the cost of a win. I assumed that the cost of 1 WAR will increase by an average of $250,000 per year, but with a standard deviation of $800,000. In plenty of years, the cost of a win will go down, even if the overall cost increases over time. There might be, say, a global pandemic, or a shortened season with no fans… wild nonsense like that.

Finally, we get down to brass tacks. When the player reaches his opt out, there’s a simple calculation. Take his new median projection, age it down as appropriate, and come up with a WAR projection for the remaining years of the contract. Multiply that number by the cost per win that we simultaneously calculated, and you have the contract that Arenado would sign, in a perfectly efficient market, if he opted out. I added one quick sanity check: if it’s within $10 million dollars of breaking even, Arenado won’t leave. That represents the uncertainty of finding a new contract, as well as the benefits of familiarity.

If there were only one opt out, our calculation would be simple. After one year, we simply apply an aging penalty and a random change in projection. Then, we use that to price out a new contract. With only one year before an opt out, things are simple like that.

In Arenado’s case, the odds are stacked in favor of him declining the opt out. The average situation (no change in projection) sees him with a $101.5 million contract after this year. He’d need to raise his 2022 projection by roughly 1.5 WAR to merit a contract that would be worth opting out for. How often does that happen? Roughly 14% of the time.

Let’s look at it more thoroughly. Here’s how much he projects to make in each scenario, including the $35 million he’s making in 2021 regardless:

In that sense, the opt out already in Arenado’s contract is “worth” $6 million. It could be worth more, though. Replace our forecasts with the ZiPS forecasts that ignore 2020 (they used 2019’s rate stats again), and instead things get a little spicy:

Hey, look! If Arenado were projected to be a bit better — and again, injury has at least something to do with why he’s not — then his option would be worth more. Neat stuff!

That’s not why you’re here, though. Or, maybe it is, but that’s not what Arenado’s contract looks like anymore. In agreeing to head to St. Louis, Arenado gained an extra opt out after the 2022 season.

It’s slightly tricky to model nested opt outs like this, where the second one can only be exercised if the first isn’t. In practice, the decision won’t be overly complex. After 2021, Arenado and his agents will model something that looks a lot like the calculations I did above, valuing his existing contract based on his current projections and including a 2022 opt out. That number will come out to something higher than $180 million, because some amount of the time, he’ll opt out after 2022 and get a raise.

With that number in hand, they’ll approximate what he could get on the open market and compare the two. If he thinks he can get more now, on the open market, than the value of his existing contract inclusive of the opt out, he should leave. Otherwise, he should stay and wait the extra year to see what happens.

Unfortunately, that doesn’t work well in my framework. In each of my 10 million simulations, computer Arenado would need to run 10 million simulations to work out the value of his option using this Monte Carlo method. That comes out to one hundred trillion simulations, and it doesn’t even add much precision for our trouble. Instead, I’m just setting the cutoff higher; Arenado will need $15 million in prospective gains to pull the trigger on opting out after year one.

How does this change things? Less than you’d think. I previously calculated that Arenado would opt out 14% of the time; he now opts out 12.4% of the time in year one. With more time to vary, and thus a higher percent chance of improving enough to opt out, he also has an 8.2% chance of opting out after 2022. All together, that looks like this:

In the scenarios where Arenado opts out after 2022, he is, on average, quite good. This makes sense, because he’s only opting out in the best 8% of scenarios. If he muddles along, he’s staying. If he has a late-career renaissance, it only stands to reason that he’ll be more valuable. All told, the two opt outs add a projected $13 million to the value of the deal — not bad!

Once more, let’s plug in a more optimistic view of Arenado and see what shakes out. This is Dan’s pre-2020 version, where he’s projected for 5 WAR in 2021:

Again, the better Arenado is now, the higher the chance he opts out, and the more the second opt out is worth to him. In the base case ZiPS projections, a second opt out added roughly $6 million to Arenado’s expected earnings. In this pre-injury case, a second opt out would add more than $17 million.

That’s the gory math of the way this complex contract works. In practice, however, Arenado seems likely to take the Clayton Kershaw route. Kershaw, too, had the choice of opting out of his contract early and becoming a free agent. He used that leverage to get the Dodgers to offer him a new contract, avoiding free agency but monetizing his ability to leave. Even if Arenado improves this year and next, that seems like the most likely eventuality.

What’s an opt out worth? In Arenado’s case, it’s real money in expectation. Add that to the extra contract year that he snagged as part of the deal, and it’s easy to see why the player’s union was okay with him taking deferrals in 2021. Remember, though: even in the scenario where he’s at his best, he still leaves less than half the time. It’s valuable despite being unlikely, and that’s the magic of options.

For years, there’s been one refrain in Anaheim: get Mike Trout some pitching help. The last time an Angels pitcher accrued 4 or more WAR was Garrett Richards in 2014, and there’s been a carousel of arms in the half-decade since. Yesterday, the Angels wholly misunderstood that refrain, sending Jahmai Jones to Baltimore in exchange for professionally cromulent starter Alex Cobb.

Cobb, who is in the last year of a four-year, $57 million contract, reached free agency after years of quiet competence. In the three intervening years, he’s alternated between being competent or hurt. He gets to it in a strange way — few strikeouts, fewer walks, and enough grounders to blot out the sun — but it adds up to something a little worse than average but significantly better than replacement level.

For the Angels, that may or may not be a meaningful upgrade to start the season. Shohei Ohtani will return in 2021, but certainly not for the whole season. When he does, he’ll likely be part of a six-man rotation. The top three starters will be Andrew Heaney, Dylan Bundy, and new acquisition José Quintana. That leaves two spots for other pitchers. Before the acquisition of Cobb, that meant Griffin Canning and Patrick Sandoval. Both are interesting, albeit unproven, options. Read the rest of this entry »

Read the rest of this entry »