Asset Valuation Part 3: Finding Bargains Cont’d.

Intro

I left you earlier in the week with a promise to dive deeper into the bargain leaderboards created by comparing projected dollar value against the expected dollar value for a pick in that draft slot. Please accept my sincerest apologies if, left without more context, you found yourself mired in the digital equivalent of a TJ Maxx aisle, paralyzed with indecision while selecting which 10 articles of clothing you want to bring to the dressing rooms.

Then, as you try on a shirt that you swear has the wrong number of buttons, or pull up a pair of pants with suspiciously shallow pockets, you may find yourself looking at the label and wondering whether these “name brands” are in fact bargains. But, on the precipice of leaving the store empty-handed but for a new batch of discount socks, you find it. The shirt of all shirts. The one which fits great, feels great, and garners compliments from strangers each time you wear it in public. That is what it’s all about.

Are we still talking about baseball? In a way, yes. Let’s look at the first batch of tables presented on Monday. As a reminder, these were the top five players whose expected bargain amount exceeds one standard deviation from the mean (in the positive direction) relative to the rest of their pick bucket.

Remember, we’ve construed the error from the pick value models as our “bargain” metric (or rip-off, I suppose, if looked at another way) here because they represent players whose projected values diverged from expectation the most. I’ve separated them by hitters and pitchers, and by league scoring setting. I omitted the 5×5 roto OBP pitcher table because it’s the same as the 5×5 roto average table. Here is what each of the columns contain:

• “Proj $” is the projected dollar value for that player based on pre-season ATC projections
• “Exp $” is the modeled expected dollar value of that pick based on the regressions I outlined in Part 3
• “Bargain $” is a subtraction of the two above
• “YTD $” is their year-to-date dollar value as of the publish time of Part 3. I haven’t updated with this week’s games so the values can stay consistent article-to-article.

Up first is the vanilla ice cream of fantasy baseball: 5×5 roto with batting average.

Roto Bargains

That there were only three hitters flagged by these criteria tells us what we probably already know: mainstream ADP is relatively well calibrated on standard roto hitters. In drafts, assuming you were picking anywhere other than 1-1, I would have stumped for the selection of all three, and I would stand by that recommendation today. It’s not our fault Aaron Judge broke his rib, and to the extent that he had any lingering reputation for being injury prone, it probably wasn’t founded given that he had played a full slate of games in four of five seasons dating back to 2021.

Meanwhile, trawling the dollar bin, you have opposite ends of the outcome spectrum. It shouldn’t have taken long to move on from Tovar, and when you did, it wouldn’t have been any skin off your back as an 18^th-rounder.

Miguel Vargas, though, is the shirt of this tortured metaphor. He was covered a few weeks ago over at FanGraphs by Michael Baumann, and hasn’t really slowed down since. He shows up on three of these four lists for hitters and is 7^th in the other, (ESPN points) still comfortably greater than one standard deviation above the mean in that category. And remember – this methodology is blind. Vargas could have been clocked as a bargain at the time of drafts based on the information available then. If you’re looking to find the value of this type of analysis, you don’t need to look beyond him. This is what a league winner looks like.

Moving on to starting pitchers, this is clearly much more of a mixed bag. Wheeler has been a big win so far, rewarding those willing to risk a pick on a pitcher whose medical affliction still effectively ends many careers. I’m still holding out hope for the other SPs in this chart – particularly Eovaldi, who simply does not have a very high bar to clear in order to justify his ADP value – but this appears demonstrably less useful than the hitter side.

This could be because I didn’t nerf pitchers when exporting from the Auction Calculator. You can do this in two ways: change the bat/pitch auction budget split, or directly decrease RP value:

I left these alone because I wanted to build models with as few assumptions from the developer (i.e., me) and rely more simply on the data. While everyone knows pitchers are more volatile than hitters, projections should be picking up useful signal on this and I preferred to let them do the majority of the work. Looking back (or ahead, at a future analysis), it might be worth experimenting with pitcher value to see how this impacts the results. In any case, moving on:

I won’t dwell too much here having said my piece about Judge and Vargas earlier. I think Soto will rebound in YTD value – if not to his preseason projection then hopefully to the ballpark of what an expected fourth overall pick was worth. That he’s missed 20% of the season to this point and still run a $24 value is impressive enough, and by the end of the year (health willing) the penalty suffered from having missed time in April will be hurting him relatively less. Wallner is another case similar to Tovar – somebody for whom you might want to spend one of your final dart throws and give a week of run before moving on. Even considering his hamstring injury, Murakami is an unmitigated success so far. While the extent to which he’ll regress remains to be seen, I would count him as another win for us thrift store shoppers.

Points Bargains

A quick note: an astute reader may have noticed very slight differences in the bargain values or order of players in these tables relative to the article from Monday. This is because I accidentally used the wrong Max Muncy’s stats in my pick value regressions. For this analysis, I updated the data. The numerical results didn’t materially change but the sorting orders for a few of these top-five lists were impacted. The largest impact is for Yahoo points above, and I have included Happ and Chapman in the table because while they were booted from the top five, they were literally the next two, six and seven.

Back to the analysis: these clusters are clearly where the most value can be found, particularly for ESPN points. In that setting, I wouldn’t be shocked to find Díaz and Arraez on most league-winning rosters by the end of the year. Even the underperforming Bregman is still returning positive value and is probably a hot streak or two from at least meeting the expectations of a pick where he was being drafted. Kwan is the only real dud from this batch, but again, he wasn’t overwhelmingly expensive up front.

As a brief aside, that is one of the things which I like most about using this analysis to build draft strategy. Many of these players so far have been free or very close to it.

This is a very low-risk, high-reward strategy. If you are in an ESPN points league, you could have reasonably chosen each of those players at their ADPs without too much positional redundancy. And if you had, your team is likely performing very well. Even where players are underperforming their projections, if you’re spending late picks on them, they usually don’t need to rebound that much for you to break even on what you paid.

Pitchers though, as with the roto categories, are more of a mixed bag. This analysis has primarily dug up relievers, and if you could have used it to nab Mason Miller or Cade Smith you’d likely be quite happy right now:

But an early pick on Diaz or Williams would have been repaid with injury, mediocrity, or both. Abreu, Uribe, and Jax (perhaps the country’s hardest throwing law firm) were essentially free and therefore worthwhile to take a shot. That’s probably how owners found Smith last year, for example. Overall though, this analysis demonstrates a clear need to think a bit more about how relievers, and pitchers generally, are treated in my pick value models.

Rip-Offs

One thing I also wanted to look at is not just whether this framework flags good, undervalued players, but also whether it can identify overvalued players to avoid at their ADPs. To this end, I looked at the 10 biggest “rip-offs” in each scoring setting. To be a rip-off here means that a player is being drafted meaningfully ahead of where their projected value would otherwise indicate.

This list doesn’t seem to be telling us much that we don’t know, which is that players at the end of drafts have serious bust risk. These results were consistent across each of the scoring settings, which is why I only included one. There are a couple of interesting nuggets here; Carson Benge seems like a genuine draft win and Jazz Chisholm Jr. is a name brand. However, relatively few players in any setting are outperforming their projected value, let alone their predicted draft value. To this end, I looked at the top five rip-offs in each category whose ADP was better than pick 200:

Now we’re talking. These lists have a few big names, including the consensus top prospect in baseball. I would hesitate to attribute his underperformance as hype given the fact that he seemed to be putting something together in May before hurting his hamstring. Likewise, Elly De La Cruz was having a dynamite year before falling victim to the injury bug, and Chourio, as noted by Jay Jaffe, has been superb since returning from a broken hand. Wood and Schlittler clearly stand as misses by this framework and warrant some future examination.

Admittedly, the results for rip-offs are less conclusive than the bargains. If I could hazard a guess though, this is due to a known blind spot of projections: young players with relatively little big-league experience. Several of these players are stars (or stars in the making) who have fewer than three full MLB seasons.

But there is some meaningful signal here, too. Each of these lists has at least one guy for whom the projections should have genuinely been heeded: Imai, Turner, Chisholm, and Simpson. Imai’s place on the list makes sense given uncertainty around how he’d adjust to MLB talent after coming across the Pacific. For the speedsters, these are all big whiffs so far for where they were being drafted. So if you’re looking at players whose ADPs are stuffed because they’re expected to steal bases, perhaps it’s best to think twice before choosing them.

Conclusion

The principle benefit you get from hitting a late pick is optionality. If you are regularly able to identify bargain players, it will open up more paths to success. Perhaps drafting Vargas has helped you weather an injury to one of the seemingly endless number of early-round stars who have missed time this year. Or if you went into the season with less draft capital after pushing in the chips in 2025, finding a few of these guys might help make up for the fact that you were missing some earlier round picks. Consistently identifying undervalued players is one of the key components to sustained, year-on-year success, and I’m hopeful that this framework will help you on your quest to achieve it.