Our Rankings vs ADP 2026-27: who we love early, who we fade

Yahoo's 2026-27 ADP switched on this week, which means the draft market finally exists. Until now, every "sleeper" and "bust" piece was a guess about what other people would do. Now we can measure it: our board on one side, the average pick of real drafters on the other, and a number in between that tells you where the surplus is.

Here's the first thing that jumped out, and it's reassuring: on the top of the draft, the market and our board mostly agree. Jokić, Wembanyama, Shai and Luka are the top four on both. Of our top 100 draft-value players, 51 are going within a dozen picks of where we rank them. Nobody is finding a first-round edge in 2026-27, and anybody selling you one is selling.

The other 49 are the article.

How to read the gaps

Two quick notes before the tables, because two of these numbers are not what they look like.

We compare ADP to our draft value, not our per-game rank. Draft value is per-game value multiplied by the games we actually expect a player to play. That matters, because a per-game rank will tell you Donte DiVincenzo is the 89th-best player in the league and forget that we expect him to play five games. Our player projections board shows both columns; the draft-day one is the one that prices in availability, so it's the one we use here.

The market here is Yahoo, as of September 16. ESPN and Fantrax drafters price things differently, sometimes by 20 picks, and if you draft there you should look at their number. Our draft edge board runs this exact comparison against whichever platform you pick, and it updates as ADP moves, which it will.

Our board is the 12-team, 9-category default. "9-cat" just means the standard nine categories (points, rebounds, assists, steals, blocks, threes, field-goal and free-throw percentage, and turnovers, where fewer is better). If your league is points, some of what follows flips, and we'll say where.

The Milwaukee reset is the biggest mispricing on the board

Milwaukee spent the summer tearing it down. Giannis and Dame are gone, Tyler Herro and Kel'el Ware came over from Miami, and the market is still pricing what's left in Milwaukee as leftovers. It isn't. It's a team with two guard spots and the center spot to give away, and four players are collecting them.

Two honest notes, because we want you to draft these players the way we actually rank them, not the way a table makes them look.

Herro is the one we'd take closest to our number. He's the alpha of a real NBA offense at 26, and two separate things keep him from going even higher on our board. We half-faded his upside, because last year's scoring came in a small sample and there's trade chatter around him, so we projected the middle rather than the ceiling. We also only give him 58 games, off the foot surgery that cost him the start of last season. That second one is the bigger haircut, and it's the more honest number: this is a second-round player, on 58 games, going in the sixth.

Ware is the one to be patient with, and it's worth being precise about why, because "we like him less than our own board does" is a cop-out unless we say what the board is missing.

The board has him 27th because a starting center who rebounds, blocks shots, shoots 52 percent and barely turns it over is worth a lot in 9-cat, full stop. All four of those are real. But look at what's underneath: that 27 is built on 14 and 10 in 25 minutes, and outside those four columns he gives you nothing. No assists, no steals, no threes, and he costs you a little at the line. It's a four-category profile, and narrow profiles are exactly what a category board overrates and what wobbles most week to week.

Now add the context. He's 22, on a team that will lose a lot of games, which means foul trouble, blowouts he doesn't finish, and a field-goal percentage propped up by a roster with no spacing. Our own note reads "a developing big, not a top-25 leap," and we stand by it. The downside version of this season is 12 and 8 on 24 minutes, and that is a round-5 center, not a round-3 one.

So pay the round-5 price. Anywhere from 45 to 60 buys you the downside case outright and hands you the board's case for nothing. That's still fifteen to thirty picks before the market takes him.

Rollins is the purest buy of the four. Milwaukee moved Dame and every initiator over 26, Herro plays off the ball as much as on it, and somebody has to bring it up 30 minutes a night.

Be honest about the room, though, because a sharp reader will do this arithmetic before we do: Rollins, Porter and Herro are all in it, and our numbers already split the ball three ways rather than handing one of them nine assists. Each is projected for six to six and a half. That's the split, and it's priced in. Even split three ways, six and a half assists with threes and steals is a round-4 player. The market has him as a lottery ticket in round 7.

Porter is a regression call, and we still like him. Herro's arrival costs him touches, our number already says so, and he's still a 117th pick who produces steals and assists like a 66th. That's the whole trick with this team: even after we mark the projections down for players sharing a room, the market marked them down further.

Buys: players the market is sleeping on

The rest of our buys, sorted by the size of the gap. Draft rank is ours; ADP is Yahoo's.

If we had to plant a flag on one name in this table, it's Derrick White. Nothing about him is exciting and everything about him is useful. He's 25th on our board because he fills five categories from a single guard slot, and 45th on Yahoo because he doesn't average 20.

One caveat we'd rather say than have you find: he shoots 41 percent, and that is a real drag on your field-goal percentage, one of the larger ones you'll take on this early. The other eight columns pay for it comfortably. Take him in round 3, know what you're buying, and let your league-mates chase points.

Vučević is the other one we'd underline. Every year the market finds a reason to let a 35-year-old center fall, and most years he plays 70-plus and rebounds like he's 28. Our games-played model actually got him wrong too: it read last season's 20-game absence as a chronic pattern and gave him 55 games, and we overrode it to 65, because one discrete injury on an otherwise durable player is one injury, not a trend. You can have him as your second center in round 10.

Fades: where the market is paying for the name

Same table, other direction. These are players going meaningfully earlier than we'd take them, and for each one we want to give you the mechanism, because "bust" is a lazy word.

LeBron is the big one, and it deserves more than a table row. He signed with Philadelphia on July 24 for two years and $8 million. That price is not a discount; it's a description of the job. He turns 42 at the end of December, and he goes from Luka's co-star, where the offense ran through him whenever Luka sat, to the fourth option behind Maxey, Brown and Embiid, with Edgecombe also needing touches. We project 30 minutes and about 16 points, down from 33 and the low 20s, and we hold the minutes up rather than cut them further only because Embiid is projected for 41 games and Philadelphia without its center is precisely when a 41-year-old point-forward still has to carry an offense. Last season's shooting was also the worst of his career. There is no 24th-season comp in NBA history, and the market is drafting him at 40 as though there were. We have him at 84. In a points league he's a safer pick than that; in 9-cat, the free-throw percentage and turnovers on lower volume make him a fade.

Jaylen Brown is the same story with a different mechanism: a base-year correction, not a decline. He isn't worse. His 2025-26 numbers were inflated by a season as Boston's only star, and his own history as a co-star, 22 to 23 points a game next to a healthy Tatum, is the honest comp. We sit deliberately above that, at 24.5, because Philadelphia's usage is thinner than Boston's was and he's the best bet to lead the team in shots on Embiid's nights off. That's still an 81st pick, not a 32nd.

Trae Young is a games-played fade, not a talent fade. Per game he's still a top-30 player. Our model originally projected 39 games, which halved a 73-game durability history over one injury-wrecked year in Washington. We overrode it to 58, because one discrete injury on a durable 27-year-old restores most of the base, and kept a real discount for the unknown in the leg. If you believe 70, he's a value at 27. We don't.

The rest of the fades are category-profile calls, and the punt-build crowd should read them differently. Claxton is a genuinely good pick if you're punting free throws. Lopez is fine if you're punting points and assists. Coby White is fine if you're building around threes. The fade is for a balanced roster, which is the default and the one most of you are drafting.

The kids: why we're lower on the young guards

This is the section that will get us yelled at, so let's show the work.

There are two different things going on in that table, and lumping them together is how these pieces usually go wrong.

One is a tax. Castle is the example. He's a real player and his assists are genuinely elite on our board, top of this group by a distance. He also turns it over 3.1 times a night and shoots 74 percent from the line, and in a category league both of those are columns your opponent banks every week.

We want to be careful here, because the easy version of this claim is false and you'd catch it: Castle is not the worst turnover line in our top 120. Eleven players are worse, including Jokić, Luka, Cade Cunningham, Giannis, Harden and Trae. That's the actual point. Those players cost you more and hand you top-ten production to pay for it. Castle hands you 18 points and 7 assists, and the market is asking for pick 56. The turnovers aren't disqualifying; the price is.

The other thing isn't a tax at all — it's that the volume isn't there yet. Harper and Boozer don't cost you anything. Harper shoots 50 percent and is a positive in turnovers; Boozer is one of the most careful players in this whole article. They're at 151 and 124 because 13 or 14 points and 3 or 4 assists in 27 minutes is a round-11 line at best, no matter how good the player is going to be. Our rookie and sophomore projections are fit on a decade of draft classes, and the pattern is boring and consistent: the counting stats arrive a year or two before the efficiency does, and both arrive later than the market is willing to wait.

The strongest argument against us is not points leagues or dynasty. It's March. Our number is a season average, October through April, and a 21-year-old's March line beats his November line. H2H leagues are decided in the playoff weeks. So when the market pays 56 for Castle, it's partly paying for the second-half version, and that isn't a stupid thing to do.

Here's why it moves our number less than you'd think. For a high-usage young guard, growth mostly means more usage — and more usage at the same percentages means more turnovers and more misses alongside the extra assists. The March version gives you more of the good columns and more of the bad ones at once. The net moves a lot less than the box score does. If you want to draft the playoff version anyway, that's fair: add a round to our number. That gets Castle to about 95. It doesn't get him to 56.

Where you should ignore us entirely: points leagues, where turnovers cost a single point and percentages don't exist, and dynasty, where you're buying the 2028 version and none of this applies. And if you're punting turnovers or free throws, Castle at 56 is a perfectly fair price — he's a natural fit for both builds, and the whole cost we just described disappears. Harper still isn't, in any build; he's just early.

In redraft 9-cat with a balanced roster, let someone else take Castle in round 5. You're taking Rollins two rounds later and getting more assists for it.

What to actually do on draft night

One thing to get straight first, because it's the mistake this kind of article usually causes: these are target picks, not our ranks. If we have someone 25th and he's going 45th, you don't take him 25th. You take him a few picks before the market would, and you bank the difference. The gap only pays if you wait for it.

• Rounds 1 to 3 (picks 1-36): draft the board, not the gaps. The market is right up top. Take the best player available and don't outthink it. The one exception is Derrick White, who's worth taking around 30 to 36 rather than letting him get to 45.
• Rounds 4 and 5 (picks 37-60): Herro at 45 to 55 and Ware at 45 to 60 are the two biggest surpluses in the article. Michael Porter Jr. belongs in this window too. If all three are gone, Naz Reid or Buzelis around 55 to 60 — though Buzelis is the one gap here that has nearly closed, so don't pay up for him.
• Round 6 (picks 61-72): Rollins. He's going 74, so you're a handful of picks early on purpose, and that's the whole cost.
• Rounds 8 to 10 (picks 85-120): the rest of the buys, each a few picks ahead of their ADP. Quickley around 90 to 99, Suggs and Vučević 100 to 110, Sheppard, Kevin Porter Jr. and Cason Wallace 105 to 115. These are the picks that decide category leagues, because everyone's first three rounds look the same.
• Let the fades go by. Somebody in your league is going to take LeBron in round 4 and Castle in round 5. Let them. The point of a fade isn't that the player is bad; it's that you get to spend the pick on someone the market underprices.

Every number in this piece is our board and Yahoo's ADP as of September 16. ADP is a snapshot, and so is a projection. It will move with every injury report and every preseason box score, and so will our board, which reprojects as news lands. The draft edge board shows the live gap for your platform, and when you're on the clock, the draft assistant runs your board live, mirrors your Yahoo, ESPN or Fantrax room as picks come in, and tells you who to take next and why. Once you've drafted, the team analyzer grades what you built.

We wrote every one of these calls down before the season with the reasoning attached, and we grade them in the spring. Some will be wrong. Ware might not develop as fast as the minutes suggest; LeBron might do the thing LeBron has done to every projection for two decades. That's the deal. You get our honest view and the work behind it, in time to use it, and you get to see the scorecard when it's over.