Where's The Edge?
Adapting My Game
Back in the day, I used to win a lot of fantasy baseball leagues. AL LABR, AL Tout a couple times, mixed Tout, Yahoo Expert Leagues, a bunch of NFBC Online ones, 12th overall in the Main Event. Going into any draft, I felt I was the favorite because I’d find the pockets of value in the auction, structure my team well, land on a few of the must-have sleepers while the rest of the nutless monkeys were overly wedded to their spreadsheets. But something changed.
My style of aggressively targeting upside players throughout the draft started to pay off less and less. I chalked this up to devoting less time to the process, and even wound up outsourcing in-season management once I realized I would never put in the time on Sundays to stay competitive in tougher leagues. Maybe it was just normal variance — yes, I had an edge, but I was playing at a breakeven level (only cashing in one out of four leagues) due to bad luck. Maybe.
But this year I’ve more earnestly gone in drafting off a hybrid cheat sheet that emphasizes projections early and ADP late. I did less research than ever, so my personal hunches/leans are restricted only to a few players, and I’m more agnostic about most. If the hybrid sheet says value, and he fits my team, draft. I don’t care if it’s a boring player my past self would never consider except at a deep discount.
I think this shift is necessary not because I’m so far removed from my prime all-in baseball years that my pattern recognition is no longer in service. I still trust my hunches on the few players who seem to me to be mispriced. It’s necessary now because baseball itself has changed, and I had not adapted to it. It started with the Moneyball A’s and then the Rays, and eventually the stat nerds won the war of who knows the most about building baseball teams. Consequently, more lineups employ platoons at different spots, fewer pitchers hit 200 innings, fewer relievers dominate the save opportunities on teams. In short, the environment around which 5 x 5 fantasy baseball was designed no longer exists.
Recognizing this, some people have tried to tailor the game toward modern baseball reality with OBP replacing AVG, for example, or adding holds in addition to saves. I have no interest in re-calibrating my valuation systems toward fringe low-stakes leagues, (though there is obviously nothing wrong with doing so if you enjoy it.) The question for me is how do I adapt to this environment mismatch between the modern game and its somewhat anachronistic derivative. The irony of fantasy baseball, played over the internet, being more traditional in some ways than the thing itself.
As much as I hate to admit it, this is where the projections nerds have it right. A bunch of 50th percentile fictions that can’t account for players sharply diverging from their prior trajectories, usually the few difference-makers whose breakouts aren’t remotely priced in — see Springer, George or Raleigh, Cal — do capture something important, and maybe it’s the only thing for the first 10 rounds or so: who is getting bankable production, barring injury or unexpected and unlikely sudden divergence to the downside.
Maybe you always had to do that for the first few rounds, but come Round 4-5 for me it was time to gamble, and by 7-10 I was swinging from the heels. The reason was sometimes those guys would hit, and you’d have a first or second-round value at less than half the cost. But even if they busted, it wasn’t that big a deal. I could find value even later in drafts and on the waiver wire to fill in. In fact, in mixed league auctions I’d just spend up for four first rounders and the top closer and fill in cheaply at the end. There were just more players with full-time at-bats and capable of 200 IP, i.e., the replacement level was much higher.
The problem now for players who are not yet fully established is it’s very likely they don’t get every day at-bats out of the gate. I had Jackson Chourio a couple years ago, and for the first two months the Brewers had him in and out of the lineup erratically before committing in June. If a prospect of Chourio’s stature can get dicked around like that, so can anyone. The teams are a bunch of tinkerers now.
But the projections take that into account. Rookies and unestablished players are not given full season’s worth of at-bats. They don’t ask “what could go right?” but “what do we know now?” And so when you draft off projections, you often wind up with a bunch of boring players like these.
I can get into how I translate projections into dollar values for the NFBC format in another post. It’s not that straightforward to model the quirks of that game, but no matter what, the raw stats will always power order in which the players appear. And from what we know, players like Seiya Suzuki, Teoscar Hernandez and Jose Altuve often show up at the top of your sheet in the middle rounds whether you like it or not.
When projectable replacement value drops, players hitting fourth (or sixth in the Dodgers’ loaded lineup) every day gain importance. If your upside swings in the ninth round bust, good luck finding 530 productive at-bats or 180 useful innings to replace them in 2026. In short, the 12-team leagues almost seem to play like 15-teamers from 10 years ago, and the 15-teamers almost like only-leagues. At-bats and innings are at a premium, floor is at a premium, ceiling is now more of a luxury than ever.
Of course, once you get out of the first 15 rounds, upside is back in play. It just takes that much longer for the pendulum to shift than it used to. The “guarantee” of quality at-bats has to become more flimsy before it’s worth risking on speculative production. But that flip does happen, and projections become even less reliable and more ignorable post Round-20. That’s where the collective value of your claims to the waiver wire starts to take on more expected return than each pick. At that point, you are buying “what could go right” tickets because “what we know” isn’t likely to matter much for purposes of winning your fantasy league.
As such, I built a hybrid system that emphasizes projections over ADP early, splits it in the middle, and goes heavy ADP late. There’s even an argument for weighting projections negatively late, meaning you downgrade anyone getting drafted just because he’s got a role for now and go all-in on sleepers and upside.
We’ll see how it works out this year, but I already feel more confident (and it’s only February) in my first two drafts after getting over the revulsion at picking the most boring and conservative players I’ve ever picked. I would say there’s zero chance for upside, but then again last year’s two biggest breakouts, Springer and Raleigh, were as boring as it got.