Asset Valuation: Part 1

“You don’t know how to play first base value assets, Scott?”

“That’s right.”

“It’s not that hard, Scott. Tell him, Wash.”

“It’s incredibly hard.”

“Hey, anything worth doing is. And we’re going to teach you.”

Introduction:

By now, 25 years hence, Moneyball concepts are likely old hat to your average RotoGraphs reader. With the suite of technical tools available to the modern fan, it can be easy to forget just how powerful those insights can still be.

Think about the decisions involved when proposing a two-for-one trade. How good is the replacement player filling the vacated roster spot? How good is the dropped player relative to the player replacing him? To know the answers, you need to know replacement value.

How do you decide when to pick a player in your draft? Or how much you’re willing to spend on one in an auction? Pricing and draft timing each hinge on expected surplus.

Or maybe you think you might be able to pay roto auction market rates for Kyle Schwarber in a points league that doesn’t penalize strikeouts. If you can, you’re most likely exploiting a market inefficiency rooted in the perceived value of real life stats that don’t matter to your league.

The real question, though, is not whether these concepts matter.

This series of articles is designed to help you, dear reader, learn how to harness them to value assets and hopefully win your leagues.

Value (ahem) Add:

Before we can get to that point though, I need to lay out some ground rules. First, I want to talk about what you stand to gain by valuing fantasy assets more precisely. While the commodification of player performance in the real game has, to some, created an undesirable amount of aesthetic homogeneity, I argue that doing so in fantasy lends richness to the experience that would not otherwise exist.

Finding an underrated contributor on the wire, nailing a late draft pick, and pulling off a great trade are the lifeblood of our game. Accurately valuing assets makes this easier. The more trades and other transactions that occur, the more owners are engaged. The more owners are engaged, the more you’re able to actually connect and build relationships with your league mates. This is what I love about competitive fantasy baseball.

With that in mind, I will take a brief detour and emphasize parsimony. Do not let the perfect get in the way of the good. The variance involved in predicting the outcome of any sport is sufficiently high that if your analysis ends up in the correct general magnitude and direction more often than not, over time, you will be successful. Outcomes obviously matter, but process does too. You probably won’t win your league every year, but usually, if your process is good, the outcome will follow ^[1]. The goal here is to give you the tools needed to start to think more rigorously than you might have before and apply the concepts to the extent you are interested and able.

To get there, you need to quantify the impact of three main aspects of your fantasy league: rules, size, and type. In the coming weeks, I will write a series of articles that details how each contributes to player valuations in fantasy. It wouldn’t be fair to leave you without a preview though, so please read on for the type of analysis which you can start to look forward to.

Practical Considerations:

For the purpose of these analyses, I’m going to reference dollar values based on z-scores ^[2]. I prefer to use z-score based rankings because that is what the FanGraphs Auction Calculator does, and that tool is simple and free. I prefer to use dollar values regardless of whether it is an auction or draft league because that is also what the FanGraphs Auction Calculator does, and that tool is simple and free ^[3].

Jokes aside, there are genuine theoretical advantages to this approach. While it might at first seem unintuitive to use the Auction Calculator as a universal tool for all leagues, it really does work just as well for valuing players in draft leagues as it does in an auction. Ultimately, the dollar value ranking is just a transformation of a z-scores ^[4], which can be used to value players in any league context.

The main alternative to a z-score based ranking system is standings gain points. This methodology is valid but requires actual standings data to calculate and only works for roto leagues. Z-scores can be applied to all types of rules settings to generate rankings. Further, I prefer converting a z-score to a dollar value because while the math underpinning them is ultimately the same, it is a lot harder for humans in society to grasp the value of 1.6543 as compared to, say, $32.

The projection system I will leverage when making comparisons is ATC. Philosophically, its “wisdom of the crowds” approach resonates with me. It also performs quite well when compared with other systems. The only material downside in using it for your analysis is that, because it relies on other systems, it only becomes available somewhat later in the offseason. If you need to run some analyses earlier in the offseason, I would reach for Steamer as it generally performs pretty well for a free system that is available basically year -round. Fortunately, however, this is not a concern for us right now because it is June.  

I will generally not distinguish between roto and head-to-head category leagues. If the categories are the same, the player values will be similar enough. The primary difference in rule mechanics will be in how pitching caps differ – but, in the spirit of what I discussed above – teasing out the difference between an IP cap of 1500 vs. a weekly start or transaction limit is probably not especially practical nor is it going to win you your league. The other difference in terms of player value relates to how their performance varies over the course of a season. This does matter, but I’m getting a bit ahead of myself here and will discuss this notion in more detail when discussing rule sets more broadly.

Now, on to the main course.

League Rules:

Rules matter. This may seem obvious, but I promise you it’s under appreciated – especially in leagues with unorthodox points schemes. Take the following stat lines:

• Player A: 2-for-3, two walks, one home run, three RBI, and one additional run
• Player B: 2-for-4, two strikeouts, two home runs, and three RBI, no additional runs

Which player is more valuable? Depends on the league you’re playing in. In a standard ESPN points league, Player A scored 12 points, and Player B scored 11. In Ottoneu FanGraphs Points, Player A scored 23.6 versus Player B’s 26. Granted, one or two points might not seem like much, but accrued over the course of a season, these edges matter. A player with a 111 wRC+ is going to be more valuable than a player with a 100 wRC+, no? Well, the same concept applies here. Player B was 11% more valuable in the Ottoneu scoring system. Knowing who to value in what context makes a big difference.

Fantasy is not the same as the real game. If you build your rankings off boilerplate lists or even your preferred WAR leaderboards, you’re not going to be valuing your players accurately. In a later article, I will show you what I mean in greater detail.

League Size:

Here is another seemingly obvious component of fantasy play. Everybody knows that as leagues become deeper, the threshold for who is rosterable changes. But how can you quantify this?

Well, consider the list of hitters that approximates “replacement value” in a standard 10-team, 5×5 league ^[5]:

These are good players! They are, however, waiver fodder in our example. Now look at the same players, but for a 15-teamer with the same scoring settings:

Please note that the order of these players in the rankings changes somewhat, most likely due to positional scarcity. These are roughly $8-10 swings. This is not an amount that is going to make or break your season, but it’s not nothing either. In a 15-teamer, J.P. Crawford has been worth as much of your budget as Jazz Chisholm Jr. has in a 10-team league. Properly evaluating players in the flabby middle of the talent distribution can seem like one of the least fun tasks an owner could undertake, but it is not any less crucial to your success than hitting some of your early round draft picks.

Allow me to present an example of an interaction of my previous two points. Now, you might look at a player like Oswald Peraza and barf ^[6] before considering adding him to your team. He is, after all, toiling on a miserable Angels team, and prior to 2026, had accrued negative fWAR across parts of four big league seasons.

And yes, in the 10-team 5×5 roto, he’s not valuable, well below the replacement level which I articulated above. But in 12-team Ottoneu FanGraphs Points, he’s been the 81st-ranked hitter ^[7]. Better than, to wit: Alex Bregman, Kyle Tucker, and the aforementioned Chisholm. Will he stay that way? Probably not, but we’re 40% into the fantasy season and banked points are banked points. Players you might scoff at in one setting (and in real life, fairly or not) might need to be rostered in every league of another and it will behoove you to know who they are.

League Type:

This is by far the aspect of fantasy valuations which contains the most nuance. When I refer to league type, I broadly mean the dual spectrums of auction/draft and keeper/non-keeper. In draft leagues, to properly value a player for your league context, you need to marry league rules, league size, the mechanism (or lack thereof) to retain players season-to-season, and what that cost is. With this information in hand, you can create dollar values for draft picks, which can be used apples-to-apples when considering trades involving both players and picks. If you introduce the ability to trade draft picks across multiple seasons, yet more layers are added. This branch of topics merits several posts of its own to fully deep dive, so I’ll keep the preview brief.

The first graph above is the expected value for each draft pick in your standard redraft, 12-team, 5×5 roto league with 23-man rosters ^[8] The second is a representation of the fitted line for each position of ADP across the sample of data I used. There are any number of insights you can draw, but I’ll leave you with a few:

• The decline in value is steepest and nonlinear in the earliest rounds (one through five). After round five, the decline is roughly linear until round 19. Subsequent picks are worth basically the same with a slight tail at the very end of the draft.
• A first round pick is worth roughly twice the value of a third rounder – an $18 difference. The difference between that 3rd round pick and a 5th round pick is likewise much smaller, only $5.
• There is a lot of noise across basically the entire spectrum of ADP ranges. If the dollar value for a given round seems low to you, this is probably why. Sure, you might expect the player you choose in the 4th round to deliver more than $16 of value, but historically, the odds that a player going in that range are going to get hurt or underperform are relatively high.
• Starting around the 18th round, picks are essentially a coin flip to provide positive value to your team. This means that picks in the 19th round and below are, on net, expected to be negative contributors to your team.

All of this information is useful. If a 7th versus an 8th round pick is the sticking point of a deal, you can look at data like this and feel more comfortable letting the higher pick go if you know they are only about one dollar different in value. Consistent, positive expected value decisions are the ticket to durable success and an analysis like this helps make them easier.

Conclusion:

This is just the tip of the iceberg. Once you know how to value the assets in your league, you can ask and answer all sorts of questions. If you have your eye on a struggling Fernando Tatis Jr. (or you want to get rid of him), you’ll have a good idea of what to offer (or accept) for your league context, because the type of offer you’d make in an 8-team redraft is surely not the same as one for a 20-team keep forever. This is the type of question every fantasy player wants to know, and I’m hopefully going to help you get there.

Post Script:

In the spirit of academic transparency, or in case you want to goof around in R for yourself, I’ll dump my code into the comment section. Here is a screenshot of the Auction Calculator setting used to export historical dollar values. ADP data are from fantasypros.com. Only double Ohtani is used. Do not be surprised if you see things in the code which I didn’t discuss or clarify above. As mentioned, I will go into greater detail in a later article.

Endnotes

[1] To this overall point, these articles will not help you build your own projections nor do I recommend trying to do so. The public, free projections available here are really quite good and you would be hard-pressed to systematically exceed their performance.

[2] Please refer to this article for a detailed but still quick rundown for how z-scores work.

[3]Please, if you are reading this series and only take one single thing away from it, let it be that you should use the Auction Calculator. It is an easy-to-use, powerful tool that will by itself put you several steps ahead of most members of public or home leagues. You can find some help using the Auction Calculator here and here.

[4] If you are interested in the math – converting z-scores to dollar values involves summing the total value units derived from your z-score based rankings and dividing the total auction budget available by that number. This creates a ratio of dollars to z-value, which is then multiplied by each player’s z-score based value to land at the dollar ranking. Truly – they are the same thing, just represented differently.

[5]“Replacement” here is roughly defined as +/- 5 spots of the 120th ranked hitter (10 teams, 10 required roster spots for hitters plus two bench bats). There certainly could be more nuance to this methodology, but this quick-and-dirty representation serves well enough to illustrate the point.

[6] Or not, given that he’s running a 125 wRC+. I digress.

[7] BRB while I go start an auction for him, actually.

[8] I will explain the assumptions underpinning the model more thoroughly in a subsequent article.