2021 First Look Fantasy Values: First Base and Second Base

November 5, 2020

Nicklaus Gaut uses the first Steamer projections for the 2021 season to project early fantasy values, comparing them to early ADP, starting with first base and second base.

The baseball offseason may officially begin with the end of the World Series but the projections offseason began last Thursday night with the release of the first Steamer projections for 2021. Created and maintained by mathematics teacher Jared Cross and his former students, Steamer remains the industry standard for non-aggregated projection systems, as well as forming the basis for many popular projection systems like RazzBall and Depth Charts.

Projections systems aren't perfect. If they were, we'd all be like Biff in Back to the Future II. Every year, different systems perform differently, in regards to accuracy and do better in some years better than others. Some do better with hitters, some with pitchers, and not just overall but within each category. Projections may not be perfect predictions (although, this guy...) but they can provide reasonable starting points, with Steamer consistently providing some of the best.

Starting today with first and second base, we're going to dig into what Steamer is projecting for this initial run, translating the numbers into fantasy dollars using a basic z-score methodology. This way, we can look at how the actual numbers compare to previous seasons (and future expectations) but also how much that projection would be worth if it actualized, in comparison to his peers.

Methodology

As mentioned, I converted  Steamer projections into fantasy dollars using a basic z-score method. As a refresher, z-scores is a popular statistical method for converting the totality of a player's scored categories, into a single number, by putting home runs, stolen bases, etc,  on a level playing field for comparison's sake. This is done by going category by category, calculating how each player's total compares to the rest of the league. Hitting 50 home runs may sound like a lot but not if everyone in the league hit at least 45 HR.

To calculate a z-score, you take the player's stat, subtract the average of the player pool, and then divide by the standard deviation of the player pool. For example, Steamer projects Freddie Freeman to hit 33 home runs, with the average of the player pool being 16.7 HR, with a standard deviation of 10.0.

(33 - 16.7)/10 = 1.6 z

A 0.0 z-score is equal to the league average, so Freeman's 1.6 z-score is saying that his 33 home runs are 1.6 standard deviations above the league average. This method is applied to each category (although the method is different for ratio stats) and each z-score is added together. This total z-score is then divided by the total z-scores in the player pool greater than zero and converted to dollars by dividing it by the total amount of fantasy dollars you have allocated for hitters or pitchers, using your chosen split.

This isn't the only valuation system and I'm not saying it's the best but it is reliable and widely used, much like Steamer. We're not trying to set these valuations and projections in stone, we just want to use them as an early and reliable touchstone as we dive into draft season. So, disagree! Whether in regards to playing time or talent rates, I'll be disagreeing plenty. But at least we'll be starting from a point of reasonableness.

Player Pool Caveats

When calculating z-scores, the player pool used is of major importance, with the pool getting watered down the bigger it gets. If you add 100 players who average 2 HR to a pool where the averages is 17 HR, the overall average is obviously going drop. For this exercise, I used a 150 PA minimum and a 12-team standard 5x5 roto league.

Feel free to quibble with the number; you should. I myself use a much more targeted player pool when making my own valuations, curated specifically for fantasy purposes. Pablo Sandoval is projected to hit seven home runs in 153 plate-appearances but do I really want him in my pool of players to be evaluated considering he has a massive negative fantasy value and won't be rostered?

But once again, I'm going for more of a neutral look in these valuations, so the player pool is set at a flat 150 PA, a total of 377 players. Just so you see how much player pool size can change valuations, moving to a 250 PA min leaves you with 305 batters (and the pool only loses eight players that were drafted in the top-430 of the 2 Early mock drafts). Ronald Acuna's top value jumps from $47 to $63; player pool matters.

The last caveat is about catchers, whom I treat as their own separate entity from the general player pool. We can get more into this when we get to them but the values for the other positions are being done without the dumpster fire of catching stats dragging everything down. This has a similar effect as limiting the pool by plate-appearances. For reference, Acuna's value with catchers in the pool is about two dollars lower than without.

Finally, a note about disagreeing and how to adjust. Projections are essentially made up of two building blocks. The playing time components (PA, IP) and the talent-rates, which is each projected stat divided by your chosen denominator (HR/PA, K/batter faced, etc). So keep in mind you might only be disagreeing with the overall value because of one or the other, not necessarily both, parts of the projection.

For example, Andres Gimenez is projected to be a below-replacement player but that's much more about his playing time than his projected talent. If he were to get 600 PA instead of his currently projected 421 PA, his line would be, 14 HR - 65 R - 64 RBI - 33 SB - .247 AVG, and his fantasy value would jump to $11.9, coming in as the #13 shortstop, just ahead of Marcus Semien.

First Base

Surprise, surprise, Freddie Freeman remains the king of the mountain. Freeman is a warm and cuddly blanket that you can wrap around your team in the first two rounds and be confident you're getting a strong base for your offense. Also, in case you missed it, Freeman basically spent 2020 treating right-handed pitching like 2020 has been treating the rest of the world.

No, seriously:

It looks like Steamer thinks that this is the Vlad we've been waiting for. After a disappointing rookie season in which the 20-year-old failed to meet the unrealistic expectations placed upon him, it turns out that Vladimir Guerrero Jr. is still probably really good at baseball. Gone was the troublesome 89.4 mph average exit velocity, jumping to 92.5 mph, and his average on fly balls and line drives went from 93.2 mph in 2019 to 94.6 mph in 2020. It obviously follows that he'd see a rise in his hard-hit rate but jumping from 38.7% HH% to above 50% puts him in a new and elite bucket of hardness.

The projection I'm most skeptical about is a .298 AVG, after a .272 AVG in 2019 and a .262 AVG in 2020. Guerrero had a .261 xBA in 2019, a .263 xBA in 2020, and wasn't getting "BABIPed" either (.308, .282). He has a premium hit tool and maybe he'll just flip a switch and start batting .300. But for now, the evidence sure seems to say .270 is more likely. Regardless, at an 80 ADP, Baby Vlad feels pretty juicy and is likely to only go lower.

I'm totally here for the continued raking of Jose Abreu but he's going to cost you on draft day. Abreu leveled up his Statcast page in 2020, posting career-highs in barrel-rate, exit-velocity, xBA, xSLG, wOBA, xwOBA, xwOBAcon, and hard-hit%. Yes, all of them. And it's not like they were bad before, Abreu was just that dominant, finishing as a top-five hitter by most valuation systems. Just looking at the RazzBall player-rater, Abreu finished at #4, essentially tied with Jose Ramirez and just a skip away from Fernando Tatis Jr. and Trea Turner.

The price is going to be high but could be worth it, even if I think he'll be a lot closer to 40 HR than the 50+ he was pacing for in 2020. There is just something to be said for the security of using a premium pick on someone that you can virtually lock-in for 650 PA and a predictable (and high) floor:

Miguel Cabrera and Joey Votto at #11 and #13? Old-timer revolution! But...Keep in mind that one thing projections don't always do a great job of is accounting for the effect of increased shifting. And both Cabrera and Votto saw big increases in 2020, although was Votto's was the most dramatic:

Looking first at their respective wOBA and xwOBA versus standard alignments and shifts, everything seems okay with Cabrera but you can already see trouble brewing with Votto:

Moving on to a better indicator of performance against shifting in wOBAcon and xwOBAcon (wOBA/xwOBA on contact only), Miggy stays level but Votto seems to have guaranteed that 90%+ shifts are coming his way again in 2021:

I have pretty low expectations for Votto, even at a 390 ADP in deeper leagues, regardless of what Steamer says. I'd rather use a late pick on upside instead of what is probably replacement-level production, at best. Cabrera, on the other hand, I don't necessarily think will get shifted more and aren't that worried if he does.

Besides the numbers above, Cabrera had a renaissance in terms of crushing lasers, posting a 93.2 mph average exit-velocity (95.6 mph FB/LD), and 49.7% hard-hit rate that were his highest marks since 2016. The problem is that asking for much more than 550 PA (even if he were healthy all season) from the soon to be 38-year-old is probably pushing it. The production will likely be useful but your league roster settings are going to determine if he's worth rostering.

Second Base

First base didn't have any monster surprises but the keystone is a different story, with Nicklaus Madrigal (yes, Nicklaus) coming in at #6. His line of 7 HR - 70 R - 60 RBI - .302 Avg doesn't look very exciting but is a great example of how using a valuation system, whether by z-score, SGP, or other, can act as an effective check against assumptions. Because regardless of if you believe in what Steamer projects, Madrigal will be incredibly valuable if he manages something near that line, even if it seems unspectacular.

The power is obviously junk (-1.0 z) and while the counting stats aren't special, Madrigal at least above-average in runs (0.6 z) and RBI (0.2 z). But being pumped up in batting average (3.1 z) and stolen bases (2.4 z) more than makes up for his other shortcoming, showing again how a high-average/speed combo continues to be more valuable than one might think. As teams keep running less, stolen bases continue to be a diminishing resource. And we've been seeing for a while that having a high batting average isn't necessarily a priority for current major leaguers.

Speaking of, that's precisely why Whit Merrifield comes in as the #1 second baseman, not posting as high of an average as Madrigal (.281, 23rd-highest) but projected with a little more speed and a lot more power and runs. He finished 2019 with 16 HR and a disappointing 20 SB (after 45 SB in 2018) but in 2020 he had 9 HR and 12 SB in just 60 games. For context, that extrapolates out to 24 HR and 32 SB in 162 games. Let's not get carried away and make him an automatic 20 HR guy but I think he can approach it and am very confident in the speed, as the Royals have shown no proclivity for stopping the green lights given to their speedsters.

Keston Hiura may have an elite 14.2% barrel-rate but his exit-velocity dropped alarmingly from 91.4 mph in 2019 to 87.4 mph in 2020 (95.1 mph to 93.7 mph on FB/LD). But what continues to be worrisome to me is his plate-discipline issues, posting a 34.6% K% (20.3% SwStr%) in 2020 after a 30.7% K% (17.5% SwStr%) in 2019. There appear to be some significant holes in his swing at the moment and that makes me wonder if a projected .253 AVG is being overly generous.

Kiura's .303 AVG was impressive in 2019 but was only backed by a .265 xBA. And this season, he had a .210 xBA along with his .212 AVG. Besides his overall average, Hiura had just a .148 AVG and .103 xBA versus offspeed pitches, and wasn't much better against breaking balls:

I'd expect that he will continue to see fewer and fewer fastballs until he can prove he should be pitched otherwise. Hiura's 70 ADP is simply way too high for a 24-year-old who's only had one-half of a good season and is coming off a Mendoza-like batting-average performance that didn't come with many silver linings.

Like a slower Madrigal, Luis Arraez is going to earn his fantasy dollars on the strength of that wicked hit-stick he carries around. Arraez had a .321 AVG in 2020, albeit in just 121 PA after missing most of September dealing with left knee tendinitis, and his .310 AVG leads all Steamer projections for 2021. That makes me love, love, love his dirt-cheap 348 ADP because the worst-case scenario is a near-guaranteed batting-average stabilizer that can be snagged late if your team needs a boost.

And saying near-guarantee might actually be a conservative estimate because his 2021 projections aren't anything new. No matter the level, Arraez hasn't ever run anything but elite averages and on-base percentages, starting when he was just 17-years-old:

If forced to bet my life (or my disappearing Nicholas Cage face, sequined pillow) on a player batting .300 in 2021, I'd put it on Arraez.

More 2020 Fantasy Baseball Advice