The Details of Our New Prospect Valuation Methodology
Today at FanGraphs, we’re introducing an updated approach to prospect valuation. You can read the announcement here, and also see the new Farm System Rankings for 2026 on The Board. This post is a detailed methodological examination of how we’ve produced our new estimates. It goes over each step of the process in order, and concludes with a sensitivity analysis. If you’re interested in the broad strokes of our new approach, the introductory post will likely suffice. But if you want to see how the sausage is made, read on.
Prospect Classes
We began with Baseball America’s annual Top 100 prospect lists for each year from 2005-2016, plus FanGraphs’ lists for 2017 and 2018. The BA lists serve as a publicly accessible bridge to the current era of FanGraphs prospect writing, and provide a nice through line with Craig Edwards’ earlier research. We took all instances of a prospect being ranked, including duplicates of the same prospect in multiple years. We converted those ordinal rankings into Future Value grades using a two-step process. First, we separated the rankings into pitchers and hitters and created two separate ordinal lists for each year. Second, we adjusted those ordinal rankings between years by a regressed factor based on that class’ major league production. This allowed us to differentiate between classes – without some type of delineation between years, every top overall hitter would receive the same grade, which is contrary to the way we grade prospects.
This method introduces some potential bias. Judging prospects based on how they turned out inherently brings some information from the future into the mix. We decided that this was the best possible way to systematically introduce varying year-over-year quality to an otherwise ordinal-only set of values, and that it also did a good job of replicating the way that grades might have actually been assigned in the past. The top pitching prospect on the 2010 list was Stephen Strasburg. The top pitching prospect on the 2011 list was Julio Teheran. It’s important to differentiate between the likely grade that they would have received. There’s some volatility in relative value assignment at the very top end of the scale based on this methodology, which is addressed in the sensitivity analysis.
Of note, each Baseball America Top 100 list contains exactly 100 prospects. We don’t limit ourselves to 100 prospects in our Top 100 anymore; this year’s preseason list, for example, had 110 players, and the number of ordinally ranked prospects with a 50 FV grade or above grew to 123 during the completion of our recent list cycle. For the entirety of this methodology section, however, any reference to the “Top 100” refers to the 100 players who were listed as the best 100 prospects in baseball in each year, to match the data we collected about past prospects.
With those adjusted ordinal rankings in hand, we assigned grades to each prospect with the constraint that the overall population of prospects matched the distribution of grades that we hand out today. In other words, we enforced the prospect team’s modern preferences on our old rankings, such that a 70 FV “means” the same thing in every year. Using this method, we produced an aggregate list of prospects with the same distribution, both of grades and the pitcher/hitter split, as our modern lists.
Team Control Value
For each player who debuted in the major leagues between 2005 and 2020, we calculated how much WAR they accrued in each of their team control years. We defined team control years based solely on days on the major league roster, ignoring actual non-tender and contract extension decisions. This results in some players having team control spans of more than seven years thanks to partial seasons, but no player had a span shorter than six seasons; if they retired before reaching free agency, zeroes were recorded for all the remaining years. We didn’t discard anyone; if a prospect never reached the majors, their ledger shows no surplus value and no WAR.
For each year of team control, we calculated an estimated equivalent free agency value using the average prevailing cost of a win in free agency from 2024 to 2026, as per this study. We used a graduated scale of dollars per WAR, in accordance with the current market. We also tried multiple other configurations of $/WAR, including a flat rate. These didn’t change outcomes by much, as detailed in the sensitivity analysis section.
We estimated cost, in 2026 dollar equivalents, for each year of team control for each player. For pre-arbitration years, we assigned the league minimum salary. For arbitration years, we assigned salaries based on a percentage of the previous season’s WAR value in free agency: 15% for Arb 1, 35% for Arb 2, 50% for Arb 3, and 75% for Arb 4. We tested multiple arbitration structures, and while none were perfect, this formulation produced the lowest mean squared error when compared to actual arbitration payouts under the new CBA. This method likely produces distortions at the extremes of player production, thanks to its use of WAR rather than the individual statistics that are used in actual arbitration hearings, but we judged it to be sufficient for a large-scale study such as this one.
With free agency equivalent value and cost for each year in hand, we discounted each year back to the present. We applied a net 7% discount rate – a 10% discount rate, in line with past research, minus an estimated 3% annual increase in the cost of a win in free agency. We discounted each year’s net value in this way, then summed them to produce each player’s surplus value. We also noted the undiscounted sum of team control WAR, and whether each player accrued two or more seasons of four or more WAR during their team control years, which we’re calling their odds of becoming a star-level player.
Evaluating Prospects with a 50 or Higher Future Value Grade
This part was easy. We grouped prospects who received a Future Value grade of 50 or higher by their position and grade. We considered every prospect in each group, regardless of whether or not they reached the major leagues. We used simple averages of surplus value and WAR, and an average of the binary yes/no star measurement to determine the odds of turning into a star. We summed up each grade and position pair and reported those results.
Evaluating Prospects with a Future Value Grade Below a 50
Given the challenges of collecting and aggregating historical prospect grades, particularly outside the Top 100, we couldn’t do a per-prospect look at the entire minors. For lower-graded prospects, we used a top-down method. We estimated the total pool of value produced by prospects outside the Top 100, then divided that pool among the grades. The logic here is that a total is far more stable than any of its constituent parts, particularly given our lack of visibility into those parts historically. We can’t say with great confidence what a single 40-FV pitcher is worth, but we can say with confidence what the entire outside-the-Top-100 group has been worth. For each historical debut class, we removed the players who were ordinally ranked in the Top 100, then tallied the surplus value, WAR, and star seasons accrued by the entire remaining population. We averaged that value across years to approximate what a typical year’s crop has been worth.
We didn’t assign all of that value to our grades below a 50 FV. First, we assigned some of it to prospects who never receive grades from FanGraphs. To approximate that percentage, we created a new measure: value accrued in the first three years of team control. We calculated the percentage of WAR accrued by prospects outside the Top 100 that was generated by prospects who never received a grade from the FanGraphs prospect team and removed that share from the total non-Top-100 WAR before assigning value to each FV tier of prospects. Second, our historical estimate of value accrued measures prospects outside the Top 100, but as noted, we assign more than 100 prospects a 50 FV or higher grade – 120 on average. Thus, we subtracted the value of those 20 50-FV players from the total value allocated to prospects graded below a 50 FV.
The remaining pool covers the average value accrued per debut class, but the minors hold many years of debuts at any given time. To account for that, we estimated how long a prospect typically remains in the minor leagues before reaching the majors or retiring – his residence time – and scaled up accordingly. We measured this for lower-graded prospects specifically, since they’re the group we care about here. We also leaned on recent data rather than the past. The minors contracted a few years ago, changing roster rules, so prospects move more quickly than they used to.
We approximated average residence time as between 1.9 and 2.5 years, with our best guess at two years even. We anchored near the low end of the range thanks to the effects of minor league contraction; measuring how long prospects stay in the minors is particularly tricky given the changing roster rules. The overall value we ascribe to the minors is sensitive to this residency adjustment, but only moderately so. Details are provided in the sensitivity analysis.
There’s one last important part of valuing lower-rated prospects, and it’s the most important part. In addition to debuting in the majors and then accruing value, lower-graded prospects can also improve while in the minors and receive a 50 FV or higher grade in a future season. For much of the player population, the odds of developing into a top prospect while still in the minors outstrip the likelihood of graduating to the majors and contributing immediately.
We valued each grade as a blend of these two possibilities. For every grade and position combination, we estimated the odds that a prospect would eventually be upgraded to a 50 FV or higher, using data from 2019-2023 (to exclude right-bounded issues). Each prospect was valued as the odds of an upgrade multiplied by our estimate of the value of that upgrade (taken from the 50 FV and above valuation method above), plus the odds of not being upgraded (one minus upgrade odds) multiplied by the expected value of that prospect contingent on not being upgraded, which we calculated up above.
An example is in order. We value a 45-FV position player at roughly $16 million in surplus value. That’s about a one-in-five chance of an upgrade – in which case he’s worth roughly what a 50 FV is – plus the four-in-five chance he isn’t, where his expected contribution is closer to $7 million. The biggest driver of a lower-graded prospect’s value isn’t their likeliest outcome – it’s the chance that they develop into a meaningfully better prospect.
Last, we split out the value of our pool across the sub-50 FV grades and positions. We based each grade’s share on the relative early-career production of players with that grade in our recent, more comprehensive rankings, the years where we’ve ranked as many players as we do today. You’ll notice that we had to make many judgment calls in apportioning value to these less-heralded prospects. Unfortunately, that’s just the reality of estimating rankings that we didn’t produce a decade ago, never mind 25 years ago. In each case, we prioritized making sure the value of the entire minor leagues was correct first. We applied some curve-smoothing to ensure that value climbs sensibly from one grade to the next rather than bouncing around on small samples. We’re most confident in the total, not the exact breakdown between tiers, but we believe they’re sensible. Further details of the range of these estimates are described in the sensitivity analysis.
Overview
By stitching together these two estimates of value – that of prospects both in the Top 100 and those who fall outside it – we produced an estimate of the value of all prospects currently ranked by the FanGraphs prospect team. The top-down construction of the overall model gives us great confidence that the top-level numbers are correct. Still, there are some inextricable limitations to this exercise. We’re using past production to make educated estimates about future production, and a great number of factors have changed in the interim. It’s not just league size – the way that drafting, scouting, and development all work now is meaningfully different than it was a generation ago. But the shape of the results – which tiers are worth what, and how the value is distributed throughout the group – holds up well across every reasonable version of our assumptions.
Sensitivity Analysis
As noted above, this analysis involved a number of estimated constants that we then applied to determine our final values. More specifically, we made assumptions about the cost of a win in free agency, the cost of a win in arbitration, the proper discounting of future cash flows, the average time a prospect spends in the minors, how likely players are to receive a higher grade, and what that higher grade is likely to be. We also chose cutoffs for measuring the distribution of results, picked different years as our sample for determining the value of top prospects, made smoothing choices, and made distributional choices about how to allocate value within the sub-50 FV tier. All of those decisions were made deliberately and represent our best estimates, but none are the only defensible choice. Rather than ask you to trust our output uncritically, we re-ran the entire model many times, changing one assumption at a time, to see how much the total value of the minors, and its distribution, changed.
The total surplus value attributable to the minor leagues is roughly $12 billion using our baseline assumptions. With the cost of a win increased by 10%, the total value of the minors would be $12.7 billion; 10% lower, and you’d get $11.3 billion, naturally. Flattening the cost of a win linearly, so that four 1-WAR players fetch the same salary as one 4-WAR player, lowers the total value of the minors to $10.5 billion, with the decrease coming disproportionately from Top 100 prospects. Changing the discount rate assumption up or down by 3% would move the total value in the minors by $700 million in either direction. Arbitration estimates don’t appear to exert a large influence on total value; we varied them by around 30%, and the resulting total value of the minors only moved about $400 million in either direction.
The total value of the minors is also sensitive to some of the calibration decisions we made. Changing the average time that low-level prospects spend in the minors before debuting by half a year moves the total surplus value in the minors by roughly $600 million. Changing assumptions about the contingent value of prospects who move into higher-value tiers could be worth as much as $500 million in aggregate, split among 40- and 45-FV players. Changing which historical years we use as inputs could be worth $200 million in either direction. We chose to start our study in 2005 to get the largest sample possible of years that roughly resemble modern baseball, and end it after 2018 because we have incomplete data on team control outcomes after that year. Other minor changes – which years to use for calibration of various distribution curves, smoothing algorithms, assumptions about the minimum salary for a negative-WAR player – mostly didn’t change the output much, affecting total minor league surplus value by less than $200 million in aggregate.
Differing assumptions made when converting legacy lists into modern-day grades changed the distribution of values ascribed to top players, even while keeping the total value in the minors intact. The published values are our best estimates, but reasonable assumptions produced outputs between $150 million and $200 million for the value of a 70-FV prospect, between $75 million and $105 million for a 65-FV hitter, and between $60 and $78 million for a 60-FV pitcher, to give you an idea of the range of values. Players with a 50 FV grade were largely unaffected by calibration; they’re the vast majority of the Top 100 dataset, and thus vary by less as calibration terms are changed. Likewise, adjustments to calibration terms for lower-graded prospects produced variation in the value of 40-FV prospects, anywhere from $3 million to $7 million of total surplus value per prospect, with that value either accruing to the 35+ or 45-FV prospects depending on the calibration type.