Trade Stephen Curry in fantasy

— -- Trade Stephen Curry.

Sell high. Hurry. Before the floorboards drop out.

Hear me out. There's every chance Curry doesn't continue to be on pace for the greatest fantasy basketball season in the history of fantasy basketball.

But deal a potential GOAT? I'd consider it.

At this juncture (the quarter pole of the season), you should know if you had what I call a "knockout draft" -- the kind of draft that, barring injury, cements you as your league's clear favorite.

Drafting Curry should have you in first place or within spitting distance of it.

But if you're rostering Curry and didn't score a draft day TKO? If you're folding Curry's rarified line (32.3 PTS, 5.1 3PT, 6.0 AST, 5.3 REB, 2.2 STL, .688 TS%) into your portfolio ... and you're not a favorite to win your league?

If you're a middle-of-the pack faux franchise with Curry? Just consider the possibilities.

Fine. Be defensive. Maybe your team is injured.

Or maybe ... your team could be butting its collective head with the collective ceiling.

Maybe you're weak up front. Maybe you blew a bundle on Curry in an auction and didn't appropriate your remaining funds with the utmost alacrity.

Selling very, very, very, very high on Curry could make a lot of sense. It could open up some interesting possibilities.

Let's look at the case, point by point.

Curry's season is unprecedented in fantasy

Steph Curry's player rater score currently stands at 24.85 points. His 2016 average player rater score -- the metric I prefer -- towers at 21.37. ( Kevin Durant is next at 17.40). That's the highest player rater average I can remember ... as in ever.

I don't remember a single player ever averaging a score that high for a full season.

To be sure, I checked with ESPN's Bradford Doolittle. His recent piece on Kobe's all-time fantasy standing included a crunching of just where players stood in the roto pantheon. He confirmed that Curry is on a record pace.

At this stage, by Bradford's chart, Curry is knocking on the door of 2004-05 Kevin Garnett.

Can Curry keep this up? The odds say "no."

You can't have the ball this much and be this efficient

Steph Curry ranks third in the NBA in usage rate at 33.4 percent (behind DeMarcus Cousins and Russell Westbrook). Understandably, Curry has the ball in his hands a lot. At the same time, he's leading the league in player efficiency rating (33.7) and true shooting percentage (68.8 percent).

Having a high usage rate is nice. It means that given the minutes, a player is guaranteed to produce a ton of volume. But combining a usage rate over 30 with a PER over 30? Again, that's unprecedented.

The highest PER in NBA history? In 1962-63, Wilt Chamberlain clocked in at 31.82. The highest of the 3-point era? Jordan hit 31.71 in 1987-88.

No high volume player has ever -- ever -- stayed this efficient for an entire season.

Curry's high PER is driven by his league-leading true shooting percentage of 68.8 percent. As I wrote last week, in the 3-point era, no player who shoots as much as Curry has ever posted a TS% that high for a full season.

It just doesn't happen. That's not to say it couldn't happen.

I'm not saying Curry starts shooting like Emmanuel Mudiay. I'm just saying a mild correction could be in store.

But take a look at Curry's last three campaigns. Something scary jumps off the screen. The most compelling part of arguing that this trend actually holds is...

This is part of the problem

In his sixth season, Curry's production has taken an Olympic-sized leap.

Curry led the NBA with 21.34 PR points in 2014-15. His current 24.84 would represent a 17-point jump.

The scary thing is that over the past three seasons, Curry has gotten 10-15 percent better every season. He's upped his usage rate and his PER, the jumps in PER coming in large chunks.

These unprecedented numbers could just be part of Curry's evolutionary pattern. But again, the odds are that this doesn't stick.

Even super-elite players can be streaky

It's hard to find a comp that fits the atmosphere Curry presently occupies. You need to find a player already playing at an MVP level that managed to improve his production by almost 20 percent.

According to our own ESPN Stats & Info, one good comp is 2005-06 Kobe.

In January of 2006 -- the month Kobe scored 81 points against the Raptors - he jumped from his seasonal averages of 35.4 PTS, 28.0 PER and 55.9 TS% to 43.4 PTS, 34.5 PER and 61.1 TS%.

From January to March 2014, Kevin Durant upped his seasonal averages of 32.0 PTS, 29.8 PER and 63.5 TS% to 34.8 PTS, 32.5 PER and 65.7 TS%.

Other examples of players registering aberrant spikes within MVP-caliber seasons include Russell Westbrook (2014-15) and Allen Iverson (2000-01).

Okay. So the odds are that Curry's numbers backpedal ever-so-slightly. But even if the numbers defy gravity and hold? If your team is middling, you should consider fielding offers for one simple reason...

You can ask for more than anyone has ever asked for in any trade in the history of fantasy basketball

If you're trading a player in the midst of an unprecedented season, you can ask for an unprecedented return.

So...what's fair here?

First off, other than for an animatronic 1962-63 Wilt Chamberlain, there's no deal for Curry straight up that makes sense. 21.37 player rater points can't be yielded from any other player on the board.

This deal shouldn't be fair. It needs to be one-sided.

So at the least, we're talking two-for-one.

Acquriring Curry has to hurt your trade partner. It has to hurt a lot. You pay a premium for premium, historic talent.

Don't forget, you'll have to drop at least one player in the deal. If you have Curry and are a middle-of-the pack team, you probably have some dead weight ( Danny Green, anybody?). But you have to factor in how much value you're dropping.

All in all, I'd expect at about a 15-20 percent markup. Say you have to drop Jamal Crawford (2.08 player rater points). You're giving up about 25 player rater points. With a Curry premium, you should be getting at around 28-30 player rater points in return.

That could mean combinations like; Kyle Lowry/ Brandon Knight, Anthony Davis/ Eric Bledsoe, or Durant/ John Wall.

Unless you're getting a combination like that -- a pair of top-12 players -- I think you have to ask for a three-for-one, or a three-for-two.

In those cases, the three players you acquire should have the feel of Harden/Batum/Butler. Or George/Walker/Millsap.

These are crazy combinations of talent. It's asking a lot. As a matter of fact, it's downright unreasonable.

Steph Curry is a luxury. If another owner can't afford it, move on to the next owner.

Trading Steph Curry doesn't make sense because his numbers don't make sense. Your asking price should conform. It should seem like too much.

But someone might bite. And the players you yield in return could turn a middling fantasy campaign into an imaginary championship.

Just consider the possibilities.