Owners Don’t Need Home Runs From First Base …
… they need Production. That’s it. If anyone says differently, they’re wrong. I’m tired of hearing owners say they want 35+ home runs from first base. It doesn’t matter where the production comes from. Owners don’t get extra points because the home run was from first base or from their shortstop. Home runs are just one category. Other hits, besides home runs, keep the AVG high and can also generate Runs, RBIs, and stolen base opportunities. Home runs don’t have monopoly on run scoring. Owners need to stop tying home runs (or any other stat) to a position and just pick the most productive players.
Today’s rant is being brought to everyone by my Twitter followers. Yesterday I asked them why Eric Hosmer was getting no love with his low NFBC ADP.
I have Eric Hosmer valued as the 28th or 32nd best hitter (z-score and SGP methods). His NFBC ADP is the 46th hitter off the board. Here are my theories for the difference. Which one do you believe is correct?
I have no idea.
— Jeff Zimmerman (@jeffwzimmerman) January 23, 2018
The big winner is power from first base. I’ve never gotten this philosophy of targeting a single stat, like stolen bases or home runs, from a set position. This is especially true early in a draft. In the first 100 picks or so, all the players are average or better. Accumulate as many of these above average talents as possible and then fill in the voids. If the team needs stolen bases, find them now later. Or batting average. Or heaven forbid, home runs.
It’s not that I don’t like home runs but an overemphasis on them can lead to a team to completely neglecting batting average. The main reason I get for ignoring batting average is that home runs will automatically create a Run and RBI. But batting average doesn’t? JMizzle stated this idea yesterday in a Twitter discussion on Eric Hosmer’s value
https://twitter.com/JMizzle36/status/955885184808865793
While some argument details are missing like the home runs for the “.310-AVG Guy” and the AVG for the “36-HR Guy”. The point to remember is that the extra hits from hitting .310 will also generate Runs and RBI’s.
Simple study time. I adjusted every hitter’s (2015-2017) HR, RBIs, and Runs to 600 PA (min 300 PA) and found how much AVG and home runs affect Runs and RBIs. The r-squared for just home runs to the combined Runs and RBI values is .47. Decent. With AVG to Runs and RBIs, it’s .18. Not so good. If I use both inputs in a linear regression, the r-square jumps to .65 (random coincidence the individual r-squares equal overall, usually the combined value will be smaller).
The equation for 600 PA works out to: 2.34 * HR + 481 * AVG -39 = Runs +RBI
To keep things simple, a difference of .010 AVG will generate close to the same number of Runs+RBI as 2 HR for a player with 600 PA. Now the mix can change depending on lineup position or surrounding talent but the totals are close.
Here is an example going back to the “.310 AVG guy” and the “36-HR guy”. Assuming “36-HR guy” hits a .240 AVG, the “.310 AVG Guy” would need to hit around 22 HR to get the same number of Runs+RBIs. There is always a balance point.
For another example, take Edwin Encarnacion (57 ADP) and Eric Hosmer (69 ADP – nice). Steamer projects “E5” at .252 AVG and 34 HR while Hosmer projects for a .290 AVG with 26 HR. So they have a difference in 8 HR which creates a multiplier of 4 (8 HR/2). Hosmer almost exactly makes up the difference in batting average from Encarnacion (.290-.252 = .038 ~ .010 * 4). Hosmer is likely to have a few more steals, so he should be getting valued higher, but he’s not.
I saw this bias against AVG with both Manny Machado and Starling Marte in my article yesterday. Remember, batting average and home runs both contribute to the other two counting stats. Owners must take each into account when valuing players.
Now back to first baseman needing to provide a team’s power. Normally, first basemen hit the most home runs and last season was no different with the position bettering right field by ~150 dongs. Owners should expect to get home runs from their first baseman. It’s almost hard not to.
Owners can go crazy if a player provides some other stat besides a position’s “traditional” stats like J.T. Realmuto’s half dozen catcher stolen bases or Brian Dozier’s middle infield home runs. They all go into the same pool.
I will continue emphasizing the concept of Production over Position to start off a draft. Get the most talented players to start before the talent begins to blend together. With no position scarcity, besides at catcher, owners can wait for the talent to level off and then fill in their team’s voids. If an owner really needs home runs, they can be found. Chris Davis and his projected 33 bombs has an ADP of 248. Why overpay for early home runs when they can be found later?
Owners shouldn’t lock themselves into just one strategy for a draft or auction. They should be ready to pounce on undervalued assets as they become available. Owners who lock themselves into needing 70 home runs or 40 stolen bases from their first two players may pass up on great values. Locking in a position plan is even dumber since every position has a range of players from great to replacement level (and worse). For those who have never value drafted, try it and compare the results. For those who have always value drafted, continue to dominate your leagues.
Jeff, one of the authors of the fantasy baseball guide,The Process, writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won four FSWA Awards including on for his Mining the News series. He's won Tout Wars three times, LABR twice, and got his first NFBC Main Event win in 2021. Follow him on Twitter @jeffwzimmerman.
Thanks for the strong article. Wondering how the math shifts in OBP, where the variance seems smaller.
Hosmer’s variance might be less with OBP but so is his performace; since 2011 when he debuted, Hosmer’s OBP ranks 98th while his batting average ranks 62nd. That said, his walk rate the past three years has been both more positive and consistent.
Great article. Just for some corroborating literature, the same trade-off in run production – two home runs is worth about ten points of batting average over a full season – appears in the Scoresheet guide.
I think part of the problem is we dont really know what we’re getting from Hosmer. 2016 Hosmer wasnt even a top 10 1b. If his BABIP regresses towards his career mark hes pretty unremarkable.
https://www.fangraphs.com/fantasy/2016-end-of-season-rankings-first-base/
This! I don’t like Hosmer, not for a lack of HR production at 1B, but because I have little faith when it comes to his production. The fact that his R+RBI is 67 lower than E5’s over the last three years is added reasoning to value him lower than E5
Interesting though, Hosmer’s net AVG difference over 3 years compared to Encarnacion is 0.085, or effectively 17 HR using the formula in Jeff’s article. Encarnacion actually hit 51 more home runs in that period than Hosmer, leaving him 34 better than Hosmer & his AVG adjustment. Using the ratio of 0.47 that’s pretty close to the 67 more R+RBI Encarnacion had. You make a good point that based on past production Encarnacion should be ranked higher, but I think this example further proves Jeff’s formula, if not give it merit.
Your hypothesis is based on a poll of your own creation in which you did not provide as a possible answer what I suspect most people would have selected — that Hosmer’s production has been wildly inconsistent from year-to-year, with a wRC+ of 113, 80, 120, 98, 124, 102 and 135. His batting average has been just as inconsistent with marks of .293, .232, .302, .270, .297, .266 and .318.
I suggest you redo the poll and include as an option his inconsistent performance.
Also, when you write that what fantasy players should look from in a first basemen is production and not home runs, I would suggest you should take a broader view. Fantasy owners also consider relative risk, which in the context of Hosmer means not only his production averaged over the years, but how much volatility there has been in his performance. A first baseman who has produced the same value, on average, over the years, but with less volatility, is likely to be a more attractive selection.
As to where one can find power while drafting or bidding on a team, it is true that one finds some power at all positions, but that does not mean that there isn’t some variance on how steep a price one pays for power based on position. In my primary league, a 12-team AL roto league with 40-man rosters, I generally find the cheapest source of power in in the outfield, and not so much because owners value power less among outfielders but because owners value speed more. Stolen bases are generally most available among middle infielders and outfielders, so at least in my league, owners pay a premium for outfielder who steal bases and hit home runs, and that leads enough owners drop out of bidding on power-only outfielders to lower prices to attractive levels.
See, now, you missed the forest for the trees on this one. Jeff’s article isn’t about Hosmer; it’s using Hosmer as an example to make a greater point: that focusing on requiring players at a specific position provide a specific stat is detrimental as a fantasy team manager. It’s like saying “I HAVE TO HAVE STEALS FROM MY SS” or something like that. We can all agree that Hosmer’s been all over the map as year-to-year production is concerned. But that’s not the point.
Jeff is actually arguing three things: Hosmer is undervalued because of his lack of home runs; first basemen don’t have to hit for power to be productive; and fantasy owners should be flexible enough not to be wedded to the notion that they must get production in specific categories from specific positions.
But the only evidence Jeff provides is an unscientific poll that lacked what would likely be the most-chosen answer, tweets associated with Hosmer and a broad comparison between Hosmer and Encarnacion that suffers from the same flaw as the poll — he fails to consider volatility and concludes that the only explanation for their different ADP is the wrongly-placed value of home runs over that of batting average. So while you and I agree Hosmer’s volatility affects his value, Jeff has not accounted for that all in presenting his evidence.
As to what you deem to be the broader argument, whether one should focus on or even consider seeking certain types of production from certain positions, Jeff really provides no evidence at all beyond Hosmer, so I will argue that a wise owner shapes his strategies to the marketplace that is his auction or draft rather than create truisms that don’t really function the same way in ever marketplace. In my main league, where owners tend to pay a premium for outfielders who accumulate homers and steals, it is generally more cost-effective to acquire a relatively higher share of stolen bases from middle infielders, because doing so enables you to get a higher proportion of power-only outfielders who in my league tend to sell for less than projected value. And because my league is OBP, that strategy is especially effective because there is always a good number of slugging, slow outfielders who hit for low-to-middling batting averages but whose many walks produce average or better OBP. If you ignored positions in my league and took up two or three slots in your middle infield with good value players who don’t produce stolen bases, whatever excess value you obtained would be dwarfed by the inflated costs of buying outfielders who both slug and steal bases.
As an example of how that works in my league, last year I had a keeper list that included a very low-priced Lindor and a low-price Andrus. I had one more middle infield slot, all of the low-to-moderate SB options had been kept by other owners as keepers, and that left me with the choice of either paying for someone who might not play or choosing among three high-priced options, of which, only one produced stolen bases, Brian Dozier. I paid slightly more for Dozier than my projected value for him (after inflation-adjustment), but I did so believing I would landing multiple outfielders who would go for much less than inflation adjusted value, which I did.
I mean, I hear ya. You don’t, per se, “NEED” home runs at 1B. But you do have to get some amount of home runs from somewhere. And it’s a heck of a lot easier to find a 30 HR first baseman than a 30 HR shortstop. If your first baseman isn’t hitting dingerz, you’ll need to make it up somewhere else. If you have Judge or Stanton, yeah you can use 1B to emphasize maybe AVG. I get your larger point – don’t automatically assume production comes from a specific position, but you do have to make up a roster at every position ya know?
I agree. I feel like using 1B for HR is kind of an easy out. And limiting. I’ll chase power elsewhere, knowing that there are usually plenty of single category sluggers at 1B available late. Or, if I need to trade for power during the season, having a weak hitting 1B that I can easily replace makes for easier deals than trying to pry away someone else’s slugging middle infielder.
You just made the argument, its EASY to find cheap 1b dingers…like from Trey Mancini or Matt Olson. Its HARD To find MI, Catcher dingers.
Why not just use the Steamer projections for R/RBI? By this measure, E5 produces about +20 over Hosmer. So if the BA/HR categories are a wash, R+RBI favors E5 (i think that’s around 1-2 SGP typically). Hosmer is +2 in steals over E5, which is usually somewhere around 0.25-0.5 SGP.
Seems like E5 having a ~1 round advantage over Hosmer in ADP is very appropriate for a player who projects to help out by about 1 point in the standings.
Further to that point, runs and rbi’s are not only a product of HRs and batting average but by team runs. On 2017, the Indians scored 120 more runs than did the Royals. That Indian team loses Carols Santana but also will have likely a full year of Bradley Zimmer, perhaps more at-bats from Michael Brantley, Lonnie Chisenhall and Jason Kipnis; I suspect run production in 2018 will be relatively close. If Hosmer returns to KC, he will likely do do without some regular contributors; Cain and Moustakas may be gone. His most rumored suitor is San Diego, which was last in MLB in runs scored.
Noteworthy: Steamer is pretty aggressive with Hosmer’s projection, forecasting him to have a career high in HRs.
Personally, I think the reason is because I perceive a very flat tier of 1B after the first 8-10. The 1B that are available after pick 150 are interesting. Carlos Santana, Matt Carpenter, Chris Davis, Greg Bird. Yonder Alonso is available after pick 300.
I tend to focus on acquiring production from traditionally weaker offensive positions earlier in the draft over more traditionally offensively focused positions. My logic is that variance between a high production shortstop and an average or even top 15 shortstop is greater than the variance between the best first baseman and the 15th best first baseman. This strategy can result in a net positive gain overall even if your shortstop is roughly equal to or not quite as good of a hitter as Hosmer. Basically it comes down to supply and demand.
Last season, 15 players had at least 30 HRs while hitting less than .260, but only 9 players had a .300 average without hitting at least 20 HRs (and 6 of those 9 had at least 11 HRs). So it is easier to pick up HRs than average late in a draft (or cheaply in an auction).
Jeff I want u in my 15 team nfbc. U can have your man crush at the 46. Ha!
A truly unique perspective – thanks for the insights!
I agree with a lot of what you write, and I understand the premise and respect that you try and and quantify it…. Buuuuuut the reason is it’s Eric Hosmer. He is all over the board in production year over year, he hasn’t seemed willing to make the changes needed to be a better fantasy contributor, and the team he is most linked too will hurt his fantasy value even more as it suppresses lh power, and is last in the league in runs. The only thing I see to like about Eric Hosmer is he is a possible swing change candidate(since he could use a swing change)…. but that is hardly enough reason to reach for someone when he has publicly eschewed sabermetrics.
“69 ADP – nice” = the real highlight of this article
Not to pile on, but Hosmer’s year-to-year inconsistency has been brought up in three well-stated comments. Quite enjoyed article as food for thought, but you’ve got a reasonable counterpoint brought up 3x. Care to address/rebut?
But isn’t one of the problems that there are (almost) no .310 guys with 22 home runs? Last year, 12 qualified hitters hit .310 or better, and only 8 of them had 21 or more home runs. Only two were first basemen: Votto and Hosmer.
By contrast, there were 16 qualified guys who hit 36 or more home runs with a .240 or better average. And 6 of them played first.
I know those are just random arbitrary cutoffs that you used, but it’s way easier to get a first basemen with power and a low average than it is to get a high average first basemen with medium power.
So if you can get Votto, obviously that’s great. Otherwise, if you don’t trust Hosmer’s consistency (which I don’t), you’re basically not getting a high average first basemen unless you are willing to settle for very little power.
On the front page of Fangraphs today I came this doozy by out own Eno “Hosmer ranked 24th on the 2009 list, then had a terrible year in the minors. After a better one in 2010, he zoomed back up to eighth in the 2011 rankings.”
Hosmer hit .241/.334/.361 over an entire season at bats with a 62.3% gb rate. Basically he was Eric Hosmer in the minors before he was Eric Hosmer in the majors. So this absurd fluctuation goes back even further. As if I needed another reason to let someone else draft him.