"I was reading Zach Lowe's article on the dangerous 1-3 pick-and-roll. He mentions that an advantage Stephen Currywill have when he and Kevin Durant run it is his ability to destroy big men who get switched on him. Now, I have seen countless highlights exemplifying this (The Stifle Tower falling in Game 1 against the Jazz). I was wondering, though, what does the data show on this? Is Steph really a giant killer?"
-- Ian Stratton
ESPN Stats & Information has been tracking defensive matchups throughout the playoffs, so I was able to sort Curry's plays this season into those against perimeter players and those against big men after switches. Curry has been very good against switches, shooting 48.6 percent from the field and averaging 1.13 points per play. But he actually has been better against perimeter players, shooting 50.7 percent and averaging 1.15 points per play.
The samples are pretty small, particularly for switches -- just 46 total plays. And I didn't separate perimeter players into those who were actually primary defenders as opposed to less-favorable defenders who picked up Curry in transition or in a scramble situation. So this is hardly definitive either way, but it's interesting to note that Curry hasn't been a giant killer in the playoffs so much as he has just destroyed everyone.
"What are the chances of seeing a comeback from 3-0 within the next 5 to 10 years?"
I don't know that I have a good way to estimate the chances, but I suspect we will see it at some point. A handful of teams have rallied from 3-0 to force a Game 7, and if you can do that, there's a decent chance of winning the series. We've also seen a number of prominent rallies from a 3-1 deficit recently -- most famously in last year's NBA Finals -- and coming back from down 3-0 isn't dramatically more difficult than that.
I do wonder whether it's possible for a team in that situation to truly have hope of coming back, given that it's never happened before. (Teams down 3-1 can at least think of examples of that comeback happening.) But I think if, for example, you had a higher seed go down 3-0 in fluky fashion or an injury that altered the series (think Rajon Rondo getting injured after Game 3 of this year's Bulls-Celtics quarterfinals series instead of Game 2), then there's nothing magical about the 3-0 comeback that prevents it from happening.
I certainly wouldn't bet on it happening within five years at even odds. Give me 10 years, and I'd consider it.
For sure. I would expect two or three of those guys in the rotation next year, depending on whether the Raptors clear a spot for a point guard via free agency/trade. If we were drafting out these players, factoring in current contracts, here's how I'd rank them.
1. Norman Powell. Clearly the best player now, at a scarce position and an incredible bargain for next season at the minimum -- though a candidate to get paid in free agency in a year.
2. Jakob Poeltl. Proved capable of playing rotation minutes in the playoffs at age 21. Biggest issue here is the abundance of capable centers in the league.
3. Delon Wright. Could capably serve as a backup point guard next season. Will he get the opportunity? Valuable rookie contract years dwindling.
4. Lucas Nogueira. Probably capable of playing backup minutes, but the fact that the coaching staff favored the younger Poeltl is a little concerning. Entering the last year of his rookie contract and not especially cheap at about $3 million.
5. Pascal Siakam. Started 38 games as a rookie but still so unmemorable that you left him off your list. Needs to either prove he can protect the rim as a center or add an outside shot to be able to play power forward.
6. Fred VanVleet. Ideal as a third point guard. Will run the offense and make a few shots, and won't hurt you while he's on the court.
7. Bruno Caboclo. Has been unspectacular in the D-League, and despite still being just 21, he hasn't shown much in the way of upside. At best, I think he turns into a 3-and-D role player. And massively overpaid for next season at $2.5 million unless he unexpectedly pushes for playing time.
From a basketball standpoint, it's not crazy, though the emergence of Jusuf Nurkic means there's no longer as much need for an interior counterweight to Damian Lillard and CJ McCollum. But from a personality standpoint, I don't think either side would be particularly interested in a reunion given the whispers about Aldridge's unhappiness in Lillard's shadow that surfaced after his departure. Aldridge also doesn't fit Portland's timetable, which is built around Lillard and McCollum.
More than that, from a trade value standpoint, it's tough for the Blazers to make moves because most of their players are either too valuable (Lillard and McCollum, and probably Nurkic too) or have too many onerous contracts (Allen Crabbe, Meyers Leonard, Evan Turner) to make sense as part of a trade. Al-Farouq Aminu and Maurice Harkless are about the extent of Portland's middle class, and unless the Spurs want to go much smaller, they don't really make sense in San Antonio. (They make even less sense for anLA Clippers team without Chris Paul.)
So I don't see a fit here, even if both sides were willing to move past the way they broke up.
"When are your final draft ratings of all the players, including international prospects, coming? Do you update your formulas based on success of players? Have you ever run formulas including measurement data like wingspan, agility speed and max reach?"
-- John Moore
I have received several questions about my draft projections. They should be out sometime during the NBA Finals, after more European leagues have finished their regular seasons. Each year, I update the ratios of how much production college and international players lose when coming to the NBA, as well as the baseline of replacement-level performance that's used to regress players' statistics to the mean. So the success of past draft picks is priced in that way.
As for measurement data, while it might have some predictive power on its own, once you have good college projections, I haven't found it to offer much additional information.