← research

Chicago traded the 4th pick for a guy who was the 4th pick

I thought the Blackhawks won this one. I checked. I was wrong, and so was everybody else.

September 27, 2026 · Keith Wilcox

On June 24, 2026 the Blackhawks sent the 4th pick in that summer's draft to Buffalo for Bowen Byram.

Byram went 4th overall in 2019.

That's the only reason I bothered looking at this trade. Usually you're comparing a prospect to a roster player and everybody just waves their hands about it, because there's no honest way to put the two on the same scale. Not here. Both sides of this deal are the same object, the fourth pick in an NHL draft. One has already been spent and we can see exactly what it bought. The other is still sitting in an envelope.

For once you can put both on the same number line.

Here's the whole deal, because the headline leaves out half of it:

Chicago gets Buffalo gets
Bowen Byram, 25, one year left at $6.25M the 4th pick
Jordan Greenway, 29, $4M, six points in forty games Louis Crevier, 25, 25 points, $900K
pick 45

Everybody said Chicago got robbed. You don't trade a top five pick for a second pair defenseman.

I thought they were all wrong. My reasoning went like this: defensemen take forever to develop, the fourth pick is no sure thing anyway, Byram is a real first pairing guy, and Chicago was drafting a defenseman at four regardless. So why wait six years for a kid when you can have the finished version now?

There are four claims buried in that and I believed all of them. The data settles three outright. The fourth turns out to be the wrong question, and the right version of it is the most interesting thing in this piece.

Short version: I end up thinking the trade was fine and everybody else ends up right that Chicago overpaid, and both of us had the wrong reasons. Then there's a fifth thing, which is the one that nags at me — the single skill Byram says he's best at is one I can't measure at all, with any data I have.

"The 4th pick isn't a guaranteed hit"

This is the one everybody leans on when they want to trade a pick, and it's the one I had most wrong.

My draft model scores every pick since 2000 from -1 to +1. Zero means the guy returned exactly what his slot is supposed to return. Positive means he beat it. A pick only gets scored once enough seasons have gone by that the answer is real instead of premature.

So how often does a top five pick actually miss?

pos_group n mean_score never_established pct_never busts pct_bust
0 D 21 -0.027 1 5 3 14
1 F 50 -0.036 3 6 4 8

About one in twenty top five picks never becomes an NHL regular.

That's not a lottery ticket. Lottery tickets mostly lose. Top five picks mostly play, and the open question is how good they get, not whether they get there at all.

Here's every defenseman taken between 3rd and 6th since 2000 who's been around long enough to judge. The whole population, best to worst:

yr pick player_name gp score
0 2008 4 Alex Pietrangelo 1087.0 0.281
1 2013 4 Seth Jones 912.0 0.271
2 2012 5 Morgan Rielly 951.0 0.264
3 2009 6 Oliver Ekman-Larsson 1137.0 0.241
4 2002 3 Jay Bouwmeester 1240.0 0.228
5 2012 6 Hampus Lindholm 829.0 0.124
6 2005 3 Jack Johnson 1228.0 0.115
7 2002 4 Joni Pitkanen 535.0 0.073
8 2002 5 Ryan Whitney 481.0 0.030
9 2011 4 Adam Larsson 1012.0 0.029
10 2008 3 Zach Bogosian 970.0 -0.002
11 2008 5 Luke Schenn 1122.0 -0.067
12 2007 5 Karl Alzner 686.0 -0.101
13 2000 4 Rostislav Klesla 659.0 -0.109
14 2010 3 Erik Gudbranson 842.0 -0.192
15 2007 4 Thomas Hickey 456.0 -0.247
16 2004 3 Cam Barker 310.0 -0.356
17 2012 4 Griffin Reinhart 37.0 -0.813

Eighteen guys. One real disaster in Griffin Reinhart, who played 37 games. Cam Barker and Thomas Hickey disappointed. Everybody else had an actual NHL career, and the top of that list is Pietrangelo, Rielly, Seth Jones, Ekman-Larsson, Bouwmeester.

So that claim is just wrong, and it was mine. The fourth pick isn't a coin flip. It's a high floor with a wide ceiling, and giving it away costs you something reliable.

One down.

"Defensemen take the longest to develop"

This one's right, and I can show you how right.

The expected value curve is fitted separately for each position and for each horizon, meaning how many years after the draft you're standing. So you can watch a draft slot pay out in slow motion and put defensemen next to forwards taken in the same spot.

pos_group D F D_as_pct_of_F
years_after_draft
1 0.126 0.110 114.5454545454545454545454545
2 0.370 0.547 67.64168190127970749542961609
3 0.837 1.161 72.09302325581395348837209302
4 1.500 1.745 85.95988538681948424068767908
5 2.153 2.514 85.64041368337311058074781225
6 2.789 3.189 87.45688303543430542489808717
7 3.525 3.755 93.87483355525965379494007989
8 4.274 4.434 96.39152007216959855660802887
9 4.879 4.927 99.02577633448345849401258372
10 4.789 5.537 86.49087953765577027271085425
11 5.223 6.609 79.02859736722650930549251021
12 5.673 6.704 84.62112171837708830548926014

Look at the first three rows.

A defenseman taken 4th gives you nothing for two years. Not much, nothing. He's still under 1.0 in year three. Forwards break away from him in year two and stay ahead for a decade, and defensemen don't get back to 85% of forward value until year seven.

Chicago's GM said the quiet part out loud when the trade landed. The odds of a 2026 draft pick helping the 2026-27 roster were, in his words, fairly unlikely.

So drafting a defenseman 4th in 2026 buys you a useful NHL player around 2030. That claim holds up.

"Byram is a true first pairing defenseman"

Now it gets interesting, because the answer flips depending on whether you measure how he's used or what happens when he's out there.

Start with the model, since he's the same pick Chicago just gave away.

The model has him at -0.018, on a scale running from -1 to +1. Byram has given back almost exactly what the fourth pick is supposed to give back. Not a steal, not a miss, the median outcome for his slot.

Which is a funny thing to say out loud about this trade. Chicago traded the fourth pick for a fourth pick that already happened, skipping the six year wait and the variance, and got the average result.

He's six seasons into a seven season maturity window, so that number can still move. He's 25. It'll probably move up.

But scoring the pick isn't scoring the player. Is he driving play? The stint table has every stretch of ice where the same twelve guys were out there, with expected goals attached, so you can just ask what happened while he was on it.

full_name season toi_min toi_rank xgf_pct xg_rank n_defensemen
109 Bowen Byram 20232024 1127 110 48.7 128 206
247 Bowen Byram 20242025 1435 42 48.0 144 212
430 Bowen Byram 20252026 1688 13 50.5 88 213

By usage he's absolutely a first pairing defenseman. 1,688 minutes at five on five last season, thirteenth most of any defenseman in the league. His coaches clearly think he's a top pair guy, and they've thought so more every year. 110th in ice time, then 42nd, then 13th.

By results he looks like a median defenseman. 50.5% of the expected goals, 88th out of 213, against a league median of 49.7%. And he did that on a team that won the Atlantic, so it isn't a good player dragged down by bad teammates.

That's where I'd have stopped if I were being lazy, and I'd have been wrong. A share is a ratio, and ratios hide things.

Split it apart and he's 34th of 213 at generating chances, 165th at preventing them. League median is 2.60 for and 2.65 against; Byram is 2.88 and 2.81.

That's not a median defenseman. That's a good offensive defenseman who leaks, and 50.5% was two very different numbers holding hands. Which, if you've watched him, is exactly what an aggressive puck-mover looks like in a box score — and a far better description of what Chicago bought.

His scoring says the same thing. A career high 42 points sounds like a lot until you notice it doesn't crack the top fifteen among defensemen. Evan Bouchard had 95. Good, not elite, which is exactly where the expected goals put him too.

So: real offensive value, real defensive cost, first pairing workload.

What I can't measure

Here's Byram describing himself the day the trade went through:

"I'm a two-way defenseman. I transport the puck. I can move the puck, whether it's passing the puck up the ice or using my feet, using my skating to transport it."

Puck transport. Retrieving it in your own end, beating the forecheck, getting it out with possession instead of flipping it off the glass and hoping.

I can't measure any of that. Not badly, not approximately. At all.

The NHL's play-by-play records ten kinds of event: faceoffs, shots, misses, blocks, hits, giveaways, takeaways, penalties, goals and stoppages. There is no puck touch, no pass, no carry, no zone entry, no retrieval. The feed tells you when something happened to the puck. It never tells you who had it.

And the gaps are enormous compared to the thing I'd want to see. Events arrive 11.5 seconds apart on average — six at the median, thirty-one at the ninetieth percentile. A zone exit takes two to four seconds. The entire play happens in the gap between two rows.

I did try a proxy: take a defensive zone event, check whether the same team's next event is up ice, call that an exit. It falls apart immediately. Depending on the starting event it fires between 12% and 37% of the time, and that next event arrives twelve to twenty-two seconds later. That isn't a zone exit. That's "this team eventually had the puck somewhere else," and hanging a player's reputation on it would be dressing up noise as analysis.

The one real signal the feed has is puck management. A giveaway in your own end is a failed retrieval, and it's recorded per player.

full_name pts takeaways_60 giveaways_60
0 Bowen Byram 42 0.72 2.75
1 Mason Lohrei 26 0.97 4.52
2 Parker Wotherspoon 30 0.97 4.39
3 Kaedan Korczak 16 0.53 4.34
4 John Carlson 60 1.64 4.31
5 Dougie Hamilton 39 0.51 4.30
6 Shakir Mukhamadullin 12 0.49 4.27
7 MacKenzie Weegar 28 1.08 4.19

Byram takes it away 0.72 times per sixty and gives it away 2.75 times. A ratio of 0.26, which is 101st of 186.

But read the rest of that list before drawing a conclusion, because it's a trap. Makar gives it away 2.77 times per sixty. Adam Fox, 3.00. Jackson LaCombe, 3.32. The giveaway leaders are the best offensive defensemen in hockey, for the obvious reason: you can't turn over a puck you never carried. Giveaways measure how much you handle it at least as much as how carefully.

Byram's giveaway rate is Makar's. It's the takeaway number, 0.72 against Slavin's 2.31, where he's genuinely low.

So the closest I can get is: he handles the puck like an offensive defenseman and doesn't win it back much. That is a long way from "here's how good he is at retrievals," and I'm not going to pretend otherwise.

This is a real limit of public hockey data, not a gap in effort. The people who can answer it either track it by hand, like Corey Sznajder's All Three Zones, or buy it from a private tracking company. The NHL's own cameras produce it and the public API doesn't expose it. Until one of those changes, the single thing Bowen Byram says he does best is invisible to me, and the honest move is to say so rather than quietly grade him on the part I can see.

Which, to be fair, is exactly why RAPM won't rescue this either. It fixes attribution, separating a player from his teammates and competition. It doesn't fix coverage. Its target is still expected goals, and expected goals has never heard of a puck retrieval.

"Chicago was drafting a defenseman at 4 anyway"

I meant this as so why not take the finished one. That's the wrong question, and the right one is where this whole thing turns around.

It isn't whether Chicago wanted a defenseman. It's what a fifth young defenseman was actually worth to this particular team.

full_name season toi_min toi_rank xgf_pct xg_rank n_defensemen
3 Ian Cole 20252026 1301 4 52.5 1 6
0 Bowen Byram 20252026 1688 1 50.5 2 6
2 Wyatt Kaiser 20252026 1314 3 43.3 3 6
1 Alex Vlasic 20252026 1415 2 42.4 4 6
5 Sam Rinzel 20252026 833 6 41.3 5 6
4 Artyom Levshunov 20252026 1042 5 41.0 6 6

There it is.

Chicago already has Artyom Levshunov, second overall in 2024. And Sam Rinzel. And Wyatt Kaiser. And Kevin Korchinski, seventh overall in 2023, a year before that. Every one of them is sitting between 41% and 43%.

That's not a shot at any of them. It's exactly what the curve said would happen. Young defensemen are bad before they're good and all four are in the hole right now.

Chicago wasn't short on young defensemen. They were drowning in them.

Spending the fourth pick on another eighteen year old blueliner buys a fifth development curve on a team that can't play the four it already has. The marginal value of one more was close to nothing. Not because picks don't matter, but because this roster had no room left for that specific thing.

Byram at 50.5% walks in as one of the two best defensemen they own. That's a sad sentence about Chicago, and it's also the trade's best argument.

So who won

Neither of us was wrong about the conclusion. We were both wrong about why, which is the more embarrassing way to be right.

The people saying Chicago paid too much were correct, and not because a top five pick is sacred. They were correct on price, and the price is worse than the headline: Chicago also gave up Louis Crevier, a functioning NHL defenseman on $900K, plus pick 45, and took on Jordan Greenway's $4M for six points in forty games. Byram has one year left at $6.25M.

I was right that the trade makes sense, and not for any of the four reasons I started with. The fourth pick isn't a coin flip, one bust in eighteen. Byram isn't an undervalued gem, the model says minus 0.018, the most average answer available. And he's a first pairing defenseman by workload, not by results.

What's actually true is narrower than either of us wanted:

The trade makes sense for Chicago specifically, because they'd already spent three high picks on defensemen who are all still in the trough. A fourth would have been redundant. Turning that pick into the one thing they didn't have, a defenseman who's already out the other side, is a real plan even if the guy they got is ordinary.

That's a roster construction argument, not a player valuation one. Most trade takes smash those two together and end up saying nothing.

So the real trade is six or seven years of cheap upside, a useful cheap defenseman and a second rounder, for one guaranteed season of a good-offence, leaky-defence blueliner on real money, with a dead contract stapled to it.

The idea is fine. They paid too much for it.

And the part I'd hold loosest is the player. Everything above grades Byram on shot quality, because shot quality is what I can see. The thing he actually says he does — retrieve the puck and carry it out — sits in the eleven and a half seconds between rows that nobody writes down. He might be better than this piece can prove. I just can't prove that either.

Draft numbers from my draft pick value model. Expected goals from my own model. On ice rates built from nhl.stints, which is every stretch of ice time where the same twelve skaters were out there.

Built from the notebook 07_byram_trade.ipynb. The models behind it live on this site: shot maps and expected goals on player pages, draft value on the draft tab.
CHIRPAnalytics

© 2026 Chirp Analytics. All rights reserved.

ThreadsInstagramTwitter