Envelope Please: 2025 crowdsourcing trade value results

Two weeks ago, we launched a new crowdsourcing trade value tool, which to respond to simple responses “Which of these two players do you prefer?” to create a comprehensive trade value ranking in our readership. With your help, we have recorded nearly 900,000 games to be exact – 897,035. Now that the Trade Value series has appeared in the books, it’s time to see how audiences line up for a wider range. Today, I will walk you through how to access and interpret your results, which can be found here and share some interesting novels about me and what I agree with or are different.

Let’s start with the exercise itself. We sampled up to 500 results from a set of submissions for each user and threw them all into a large number of showdowns. We ordered these showdowns randomly and then used ELO ratings to turn the match into an ordered list. We will then order randomly 100 times and average the results, which reduces Elo’s bias against recent matches. This creates a total preference list of the crowd. The first thing you see when you open the above link is the complete result. For example, I’m very close to matching my official list:

This page will show you your list (naturally, my list is called “Ben Clemens’ Trade Value Results”, but your username will have your username). You can also see players who you like the most and dislike the crowd. Due to the way we designed this tool, the list might get a little weird in the middle (less in the match we give you), so if Maikel Garcia ends up in your top 30:

If you like the tier system I used in the last two years of the Trade Value series, you are in luck. The Graphics option turns your ranking into a Z-score and then use these options to make layers. If you hover over the face on the player on the tool, you can see their ELO scores based on the matchup you selected and their numerical ranking. They will also be highlighted in the crowdsourcing layer on the right so that you can compare your work with the work of the group. These work is based solely on these Z-scores; everyone with similar scores will be grouped together. If, like me, one player is much taller than the rest of the group, you can see this clearly:

Finally, you can view the full crowdsourcing results, as well as a full breakdown of my list compared to the crowdsourcing list in the tag on the right. As mentioned above, a single list in the middle may become a bit trendy, but the total crowdsourcing list is a large enough sample that the effect disappears. One thing I absolutely love to see is that my rankings and crowdsourcing ranks nine out of the top 10 (I have Cal Raleigh, and you have Pete Crow-Armstrong).

Now, let’s take a peek at some interesting tidbits that emerge from this exercise, starting with my list of top players in the crowd:

Boys, you will definitely hate catchers! I want to come up with some extensive generalization here, and that’s what stands out. Will Smith has little to do with one of the top 50 trade values for baseball in a baseball game, only you don’t value catchers at all, but that’s definitely in line with the rankings of the other crowds. I think I’m already very low on Raleigh (ninth) and he’s the only catcher we’re close to (11th in the crowd ranking). I went into detail about how I tried to change the catcher’s valuation, but obviously the crowd is Distant Lower people wear ignorance tools. Similarly, if a player is injured recently, the crowd doesn’t like it. I think this makes sense. It shows different preferences, but this seems to be good. It also makes sense to bring Spencer Strider, Eury Pérez and Yordan Alvarez to the top 30 and get to the bottom of the top 50 as they are currently injured or injured, even though I chose a different option.

After that, I think you’ll see some general disagreement on the stars that sign big contracts. Corey Seager? It is greatly unpopular among the crowd. The same goes for Zack Wheeler, which surprised me, Byron Buxton’s contract and injury status are difficult to resolve. Then there were only a few scatters: two pitchers, who were commanding at Logan Webb and George Kirby’s West Coast, and then strongly turned down the crowd on cheap team control at Jordan Westburg. I have a high consensus in Westburg, but I think it’s arguably much lower crowds. After all, I showed this list to a group of people at the front desk who didn’t tell me that Westburg should be out of the top 100.

The plot gets thicker as we look at my roster of players below the crowd:

If I had to summarize here, I would say there were more crowds than me. Roman Anthony ranked in the top 20, Wyatt Langford and Jacob Wilson ranked in the top 30; these are positive valuations of young players with short-circuit records. I wrote extensively about how I think about the way these guys value in every mess they are, but the crowd obviously disagrees. I can see quite a lot of devaluation and chase rate devaluation here: a large number of shortstops and outfielders are more popular among our readers than me. Checking this list tells me that the crowd prefers potential more than I do, which is not surprising. The potential is really fun! For me, one of the hardest parts of this exercise is reconciling the fact that the purpose of baseball is to win games with an unshakable feeling that newcomers are always better. The easiest explanation I can provide for this difference list is that we deal with this fact and make a difference. That said, “Jacob Wilson’s taller, Jordan Westburg’s position is much higher” is a tough position, and I think that’s largely thanks to Wilson being the updated, shimmering, more batting average option.

Finally, here are the players with the highest crowd, they didn’t make it to my top 50:

Most of my “just missed” queues are there. Mackenzie Gore, Jeremy Peña, Jackson Holliday, Nick Kurtz, Cam Smith and Andy Pages, all of which would be in my top 60 if I ranked high. Likewise, all of these players except Juan Soto are in the second half of the top 50 in the crowd, where relative valuations are compressed. No. 35 Blood is different from my ranking (No. 51), but not a ton. If I ended up with these players in the final rankings, it wouldn’t be strange. Actually, the list of people looks good to me. However, I will admit to being a little confused about some of the inclusions in this particular list. Soto is a great player. It’s weird to get him 24th and then hit Seager outside the top 100. Similarly, Spencer Schwellenbach and Cole Ragans are now both in 60-day IL. The idea I would like to be that despite being hurt, you can put them on the list, but I have a hard time coming up with a through line while pushing these guys on the list, while the Strider/Pérez/Alvarez Group is down. I’m not saying this is impossible, it’s just that the overall preference of the crowd can make some weird groupings.

I also bet that if we do this exercise again today, Kurtz will be higher. Most of these votes happened in the first few days. Kurtz had 138 WRC+ in the All-Star Game, which was when the tool debuted. By the end of last Friday’s vote, eight games, the 472 WRC+ Stretch had a 184 WRC+ season roster. I estimate his time-adjusted position in the mid-20s, but it is certainly hard to say. Those terrible players, keep playing as we try to rank them!

Basically that’s it. You’ve done the hard work of ranking all of these players; now you can do whatever you want. I even have a little bonus stat: the five closest matches between the top ranked players. We directed our algorithm to have more showdowns between the most important players in a single list so that we have more data about who each user prefers in the best player in baseball, which means more showdowns. But even with all the extra volume, these options separate in the middle:

Thanks again for your help to make this new tool useful – we can’t do it without everyone. Of course, I would like to offer Keaton Arneson and Sean Dolinar a round of thanks for designing and developing crowdsourcing tools and results pages. Hope you enjoy wearing my shoes for a week.