Asset Valuation Part 2: League Context

Have you ever looked at art that changed based on the direction you’re viewing it? Or maybe stared at that famous drawing of a duck for 10 minutes, only to have it morph permanently into a rabbit? The notion that one thing can be two or more does a number on our brains sometimes.

And while I don’t have any data to back it up directly, I think it’s fair to say that the majority of fantasy analysis focuses on 5×5 roto. Even I am guilty of it. All too frequently, the industry overlooks the fact that fantasy performance is differently valuable based on how you look at it.

This leaves a decent chunk of fantasy players in the lurch. If you play in multiple leagues, it is likely that they don’t all have the same scoring format. Maybe you are in a competitive 5×5 roto you found via classifieds on Reddit, but also have a home league with a convoluted point system because your 8^th-grade buddy thought it would be fun to make triples worth 15 points. The value differences across these types of leagues can be meaningful. And while a rankings article might make passing reference in the blurb for Kyle Schwarber that he’s more valuable in OBP leagues, how can we figure out the extent to which this is the case? Below, I am going to walk you through how the FanGraphs Auction Calculator can be used to elucidate the extent to which player values can differ based on league rules.

In my analysis, I’m focusing on four different types of leagues: 5×5 roto with batting average, 5×5 roto with OBP (no batting average), ESPN points and Yahoo points. I have chosen this tasting menu of settings because they each highlight a key aspect of how valuations can differ across leagues. The OBP versus batting average distinction is self-explanatory. Further, the roto settings should value players who steal more than points settings would. Meanwhile, ESPN points penalizes strikeouts, while Yahoo points does not. I am using a baseline league structure of 12 team, 23-man rosters:

But for your own, feel free to enter whatever applies to your leagues. I’m pulling 2026 year-to-date data from the Auction Calculator for each setting. If you want to see projections, simply change the setting dropdown to your preferred system. Be sure to select the various “Rest of Season” options because if you don’t, the projections you get will be as of the start of the season.

Unlike some of the other work I’ve published so far, this analysis can be done entirely in Excel and does not need any programming knowledge. The first thing you’ll want to do is download the exports based on each league setting. For the purposes of this analysis, I recommend deleting columns that are not the player name and overall dollar value. You will need to export and combine pitcher and hitter datasets separately; this can easily be done via copy-and-paste.

However you condense the exports is fine – whether you want to aggregate all hitter files into one and then combine with an aggregated pitcher file, or aggregate hitters and pitchers for each league setting and then combine that way. As long as you are consistent about how your cells are sorted your final sheet should be the same. Your end result will need to be one sheet with n+1 columns, where n equals the number of league types you have exported and the extra is for player name.

With this, you now have a surprisingly powerful (yet still simple) tool at your disposal. It is a Choose Your Own Adventure from this point regarding how you want to analyze your data. If you are in only two leagues, you can start as simply as subtracting your two value columns and sorting the sheet by the difference. The players at the very top and very bottom of the list are guys whom you should be mindful of when considering them in transactions in your leagues.

Likewise, as your number of league settings increases, you can add complexity to how you measure the differences. In the analysis below, I have used standard deviations (still doable in Excel!) in a few different ways to determine which players have the highest variances based on league setting.

The first table is based on raw standard deviation calculations, and the second is based on coefficient of variation, which is the standard deviation divided by the mean. The latter will allow players who are lower valued but still highly variable to be identified. For hitters ranked by their coefficient of variation, I filtered for players who had at least one category greater than $10. Without this nuance, the leaderboard is dominated by players whose mean value is a decimal which creates a few astronomically high but analytically uninteresting results.

In the raw dollar value table, we have an unsurprising crew. Four players who strike out a lot, and one who never does. It should come as no shock that ESPN Points penalizes strikeouts, and the effects are stark. And while we may be seeing the Oneil Cruz breakout we’ve all been waiting for, Luis Arraez has outperformed him, and Kyle Schwarber for that matter, by a significant margin in that setting. And speaking of Schwarber – in OBP roto, he’s been 28% more valuable than AVG roto. In Yahoo points versus ESPN, he’s been more than twice (!) as good.

For players sorted by coefficient of variation, there is less name value. This is intuitive because the denominator in that calculation is mean value, and smaller denominators (i.e., smaller mean values) will create larger coefficient ratings. There are still, however, some pearls to pull from this oyster. Willy Adames’ topline stats might not look great, but he’s been a solid player in Yahoo points. He might present a genuine buy low opportunity – his ADP was north of 130, reflecting what is likely to be a fairly low initial investment, and whoever is rostering him is probably low on patience. Nothing under the hood suggests a glaring problem, so if you can get him for a relatively cheaper cost, take a stab at it.

Now, let’s take a look at pitchers, with the same filtering rule applied on the second table. I have also omitted the OBP roto column because pitcher scoring was the same for each and while there was some small variance in the dollar value (~$1 generally), leaving them in essentially double counts the roto categories in the variance calculations.

Christopher Sánchez is obviously a top arm anywhere, and Gavin Williams appears to have found a new gear and is likewise performing well regardless of setting. But these guys’ knack for going deep in games and get wins adds an extra value in ESPN points. Meanwhile, Cease and McLean (with but six wins between them) are buoyed in points relative to roto by their respective penchants for punch outs.

If you’re looking to make a trade based on this data, good luck prying Sánchez away from anyone. But Cease, Williams, or Connelly Early in points? Comparatively more doable. These types of moves aren’t especially splashy, but they are the type that, if made consistently and in the right leagues, build winners.

Next up is a list of the top players ranked by average value across the settings types, but with their ranks in each respective settings compared.

First of all, congratulations to the Jacob Misiorowski drafters among us. One could easily make the argument that he’s been the best player across all of fantasy so far. Second, while the analysis presented above is not revelatory (good players are good players), there are some things to learn, like knowing that Oneil Cruz has been just the 76^th-best player in your ESPN points league when he’s untouchable everywhere else. Or that Nick Kurtz, while good everywhere, has been a top-three bat OBP leagues. To that end, here are two lists of players based on the largest standard deviation of their relative ranking. Pitchers and hitters are separated, given the issue with roto categories described above.

For hitters, again, the list is unsurprising; you have guys who strike out a lot and guys who don’t. Well, and Taylor Ward, whose unorthodox production (recently highlighted under our green banner by Ben Clemens) dings him in roto but not points.

Frankly, I was expecting stolen bases to play a larger role in driving variance than strikeouts or lack thereof. In other words, I was expecting to find more guys whose value spiked in roto settings relative to points. Usually, though, it was the other way around in my data. This is worth poking into further – perhaps it is just an artifact of my study setup – but it could point to a notion that players who derive their only fantasy value from steals are going the way of the dodo.

For pitchers, this list demonstrates that some arms can be both borderline unrosterable in roto (and thus ignored in the majority of mainstream fantasy analysis) but also be genuine impact players in points and are therefore worth your time to know. The vase between the heads, as it were.

The analysis presented above, while straightforward, is a building block to other aspects of valuation in fantasy, especially in a draft context. Looking to hunt for surplus value in a draft? You’re going to want to know how a player fares in your league settings. Or maybe you’re building a list for the opposite – players who are overvalued for their ADP and you want to avoid them. Again, you’ll need to know how they rate in the context of your league. In my next Asset Valuation article, I’ll show you how this works.