After downloading nearly a hundred free rankings from various experts and sites like Sleeper, Yahoo, ESPN, CBS, and NFL.com, I created a new methodology to evaluate which preseason projections actually translate into success.
The analysis involved using an exponential decay model to convert rankings into values, emphasizes the importance of getting those first few rounds correct since early-round picks significantly impact your final results.
I then scraped thousands of fantasy leagues to determine each team’s placement, ranked by PPG instead of wins to decrease variance, and had each set of rankings choose the most valuable team they think would dominate.
What makes this approach better? It replicates real-world scenarios where cross-positional evaluate matters just as much as individual position performance, necessary for building a competitive roster.
The tested sets included ADP from major platforms, 78 individual posted rankings online, FantasyPros Expert Consensus ECR, FantasyCalc.com trade values generated through an algorithm on millions of real trades scraped from FleaFlicker and MFL, plus week 5 rest of season projections to compare vs preseason predictions.
Why does this methodology matter? Because it provides an accurate representation of how well rankings performed when used to draft teams, heavily considers detail that most analysis misses, and scored how accurately the most valuable preseason teams actually performed, giving you a clear answer about which rankings result in better performance across multiple league types and rounds.
ADP Analysis
When I found ADPs from major fantasy sites like Sleeper, CBS, and NFL.com from August 17th, 2023, I used the same date for FantasyCalc rankings to perform a fair comparison, and the graph shows something fascinating: the best preseason team ranked by FantasyCalc values was actually the best team ~19% of the time, the 2nd best ~14% of the time, and finished in the top 3 more often than teams ranked by ADP values, while being less likely to finish in the bottom half.
Using this methodology, trade values performed better than all ADP sources, with Sleeper ADP next best, followed by CBS and NFL.com. Both are user decision based, but I hypothesize FantasyCalc values perform better because users are more informed when making trades versus draft picks.
Trade Analyzer Fantasy Football
In a draft, you have 1 to 2 minutes to decide between 5 to 10 players, you don’t know your entire team yet, and you might not even pick the player you think is most valuable if you think they will fall or there will be a run on a different position soon.
In a trade, you know your entire team’s strengths and weaknesses and can better assess cross positional value, you have unlimited time to research the 1 to 3 players you plan on trading or trading for, and can more accurately calculate player value.
Expert Rankings Analysis
After I scraped every free set of 2023 expert draft rankings I could find, which totaled 78 different experts, the majority of these were from 9/7, which is what I used for the FantasyCalc values comparison, and the results were surprisingly eye opening: FantasyCalc trade values performed 7th of the 78 rankings, landing in the 91st percentile and proving that rankings generated from trade data were surprisingly effective and able to outperform the vast majority of expert rankings.
A few experts unsurprisingly beat FantasyCalc values if they correctly bet on specific sleepers or against certain top players, but here’s what I find fascinating: there are likely some rankers that use datasets to help their rankings which could contain more accurate signals than what can be captured by trade values alone.
Tips Of Trade Analyzer Fantasy Football
Why does this matter for your draft? Because algorithm generated rankings can consistently deliver top tier results without the human bias that some experts bring, and they’re more reliable than the vast majority of individual expert opinions you’ll find online.
FantasyPros Expert Consensus Rankings (ECR) Analysis
FantasyPros slightly beats FantasyCalc’s trade values and landed 3rd place of 78 expert rankings, showing the wisdom of the crowd works since ECR similarly benefits from pooling multiple opinions, but here’s what I discovered: FantasyPros is one of the few sources that also provides rest of season rankings throughout the entire season, and when I happened to download their week 4 rest of season rankings to compare against FantasyCalc’s, the results flipped dramatically because FantasyCalc greatly outperforms ECR once the season starts.
With months of prep time before kickoff, it makes sense that preseason average expert rankings are better than what the preseason average trade’s ranking shows, but once games begin and player value begins to shift rapidly, it’s incredibly difficult to rank 200+ players by hand every week.
The average expert essentially faces a huge data disadvantage when compared to hundreds of thousands of real trades happening across platforms.
Why do trade values outperform ADP when both are user decisions? Because users are more informed when making trades vs deciding between 5 to 10 players to draft in under 2 minute timeframe windows.
FantasyCalc trade values outperform 91% of all expert rankings, and while ECR outperforms trade values during preseason since experts with months of prep will naturally outperform the values from the average trader, however, once the season starts, trade values greatly outperform expert ADP because live data crushes outdated projections when deciding rankings.