The Diff: Kyrie Irving’s potential

Today marks the start of a new era: The Diff, where I’ll write lots and lots about sports stats. Kudos to friend and fantasy football semifinalist Marc Sugerman for coming up with the article name, a shoutout to the Q’s scoreboard. And thanks as always to Mike Brenkus for the design.

The Diff

For my first-ever edition of The Diff, I wanted to focus on some thoughts I had about Kyrie Irving over the last week or so. This all started after the Cavs’ impressive 99-83 victory over the Hawks a week ago today. I was at that game with my dad. But the following conversation took place the next day on Gchat with my brother Sam, who you all might recall from our combined offseason post back in June. Hope you enjoy.

Sam: We do have the hardest part of rebuilding out of the way: We have a superstar. If Kyrie isn’t a top 10 player in the NBA already, then I don’t know who is.

Jacob: That’s the argument I started on Twitter right now. In Rick’s recap from the Hawks game, he wrote that Kyrie clearly has the potential to be a top 5 NBA player. It’s an unwinnable argument, as Rick said, because we’re dealing with “potential,” but for some reason, I think I disagree with that. I also clearly don’t think he’s top 10 right now. That’s a huge reach.

Sam: Who are the top 10 then?

Jacob: Let’s just do a quick comparison. Is Kyrie better than James Harden?

Sam: Maybe not today but in about 3 years, I think there will be no doubt. They are close.

Jacob: Close yes, but I’d probably give the slight nod to Harden.

Sam: But Harden has about 5 years on him. Kyrie is what, 20? When Harden was 20 he could only dream of playing how Kyrie is today.

Jacob: Harden has 2.5 years on him. Harden: 8/26/89 and Irving: 3/23/92.

Sam: OK, so at 20 years old, the only comparison might be LeBron. Kyrie had a better rookie PER than LeBron did:

Jacob: That link is a search for all players 21 or younger on Feb. 1 (Basketball-Reference’s stop point) that have had 20+ PPG, 20+ PER and at least 25 games in that season. Only two players have done this in the last seven years: Kevin Durant ’08-’09 and ’09-’10, and Blake Griffin ’10-’11. Kyrie is on pace for that this season.

Sam: Also, if starting a franchise today, name 10 players you’d take over Kyrie. I couldn’t.

Jacob: This is for all player seasons ages 22 or younger, height 6-5 or under, PER 20+: One name shines through: Chris Paul. Chris Paul’s age 20 season PER: 22.1. Kyrie Irving’s age 20 season PER: 21.5.

Sam: And Chris Paul as of today is the MVP. I think Kyrie will be an MVP candidate in about 3-4 years, if the Cavs or whatever team he may be on is winning. I think Kyrie’s PER is going to rise this season too. He is gonna have a monster 2nd half. I think it finishes around 23.

Jacob: It’s because his usage % is so outrageously high as well. 30% is so rare for a kid his age, and that’s just going to go higher without Andy now. PER is not a pure rate statistic. The more points you have, the better you’ll do. It’s not necessarily just based on how efficient you score each point. Here’s now a search for all age 21 or under, usage rate 28%+ and minutes played 750+: You can see this just doesn’t happen very often.

Sam: What is usage rate?

Jacob: Per Basketball-Reference, “USG% — Usage Percentage; an estimate of the percentage of team plays used by a player while he was on the floor.” Practically, a metric that measures a player’s involvement in the offense. Kyrie is squarely at 30.0% this season.

Sam: So, if one has a high usage rate vs someone with similar stats, what does that mean?

Jacob: For example, Tristan’s usage rate this season is 16.0%. Omri’s is 18.2%. Harden’s is 29.0%. LeBron’s career avg is 31.7%. My point technically is that Kyrie’s high usage rate potentially upwardly influences his high PER. I’d have to get a whole lot of data to prove that correlation. … But it’s a hunch. Nonetheless, it’s not like he’s not efficient per point and per his usage. He’s as efficient as Chris Paul.

Sam: How good were the Hornets in Paul’s 1st-3rd seasons? Were they a playoff team by year 3?

Jacob: Intriguing. Chris Paul’s usage rate in his first two seasons: 22.6%. His PER in those 142 games: 22.0. Kyrie’s usage rate in his first two seasons: 29.2%. His PER in those 80 games: 21.4.

Sam: Dang, Hornets got the #2 seed and made the 2nd round of the playoffs in Paul’s 3rd year. Hard to see the Cavs doing that.

Jacob: Paul’s third season: 80 games, 25.7% usage, 28.3 PER. Wow.

Sam: He became a stud in year 3 and led the Hornets to the 2nd seed. I dunno if Kyrie and the Cavs will make that leap just because of the talent around him.

Jacob: Yeah, obviously Paul had much, much more talent around him. But that’s unbelievable for someone to have that good of a PER with that low of a usage rate, in my opinion. A peculiar statistical relationship, but I’m so intrigued about it right now. So here’s where I’m at: Kyrie Irving projects to be Chris Paul-lite. Thoughts?

Sam: I agree. Paul is the best PG alive.

Jacob: Paul is on pace to be one of top 3 PGs ever.

Kyrie Irving is seeing plenty of extra defensive attention this seasonSam: If Kyrie can become 90% of Chris Paul, I’m happy. I’ve always thought that’s who he resembles most. They have different styles, and I think Kyrie is a better jump shooter at this age. But Paul is a wizard on the court when setting up his teammates and also defensively.

Jacob: Paul’s third season: 80 games, 25.7% usage, 28.3 PER. Paul’s fourth season: 78 games, 27.5% usage, 30.0 PER. So good. Paul is leaps and bounds better defensively, as our friend Mitchell tweeted out today too. And yeah, I agree that Kyrie is a better pure scorer at this age. Paul’s steal % in those third-fourth seasons also is insane: 3.9%. That is the percentage of opponent possessions that end in a Paul steal.

Sam: Yeah his steals were nuts.

Jacob: For comparison, Kyrie’s career steal % is 2.1. LeBron’s is 2.3.

Sam: Paul would have games with 5 or more steals all the time. Even last week he was insane with those steals vs Dallas to steal the win. He’s so competitive too. I think Kyrie shows that fire, but not quite where Paul is. We need to start winning.

Jacob: Yeah it’s ridiculous. Looking at historical comparisons, lots of guys had 4.0+% steal rates in the ’70s and ’80s. In the modern era, only Rondo/Payton were as good with steals at that young of a age as Paul (and, intriguingly enough, Eric Bledsoe, in his limited career minutes thus far).

Jacob: Found another great stat. So here’s what I searched: guards in age 21 or under season, with 24%+ usage rate, 750+ minutes played, 55%+ true shooting. Kyrie’s ’12-’13 stats: 30.0% usage, 1033 minutes, 56.3% true shooting. Three players ever: Jordan, Magic, Kyrie.


Now for some closing thoughts to bring this all together.

In the NBA, my strong opinion is that forwards dominate. You’ll find this in PER and just about every all-encompassing NBA metric because of the importance of rebounds. Guards also have had a historic tendency to be less efficient in shooting (although Kyrie certainly is the perfect picture of an efficient scoring guard), especially compared to big centers that can shoot close to 60% near the basket. Among the 26 eligible players with PERs of 20+ this season, only 8 are guards: You could find a similar trend in other stats.

So here’s my line of assumption following that PER pattern: In order to be a top 25 player (i.e. All-Star-level), Kyrie, approximately, would probably need to be a top 10 guard in the NBA. In order to be a top 10 player, Kyrie would need to to be a top 5 guard. In order to be a top 5 player, he’d need to be a top 2 or 3 guard in the NBA.

Since the start of last season, Kyrie ranks No. 8 in guards at PER at 21.4: This season alone, he ranks No. 7 in guards at PER at 21.5: (The lone difference is Manu Ginobili, who drops from No. 5 to No. 8 in these two comparisons.)

But, in these PER rankings, Kyrie ranks ahead of at least two guys you probably wouldn’t expect: Rajon Rondo and Deron Williams. The players ahead of him are almost to be expected: Paul, Kobe, Westbrook, Harden, Tony Parker and Dwyane Wade. So certainly, according to PER, Kyrie is in the debate somewhere in the top 10 among guards. That seems pretty clear. And that’s why he’s on the cusp of the All-Star debate this season.

On the flip side, however, if one were to look at Win Shares/48 minutes — another all-encompassing statistic that has its positives and negatives — Irving ranks No. 22 among guards both for the last two seasons and this year. WS/48 is broken down between Offensive Win Shares and Defensive Win Shares. Kyrie ranks horribly in Defensive Win Shares, and thus falls way down in the rankings. It’s odd he’s so low, but you have to have this statistic in any overall consideration set.

So Mitchell’s point is valid about Kyrie being so awful defensively right now, that how could he possibly be a top 5 player (as in, top 2 or 3 guard) in the NBA ever? But then you wrestle back to this final point: Kyrie is 20. He only turns 21 in March. He’s younger than Dion Waiters. In fact, Irving is the fifth-youngest player in the NBA when you look at players with at least 750 minutes played this season:

That’s why folks are so positive about Kyrie and his future. It’s why outlets such as ESPN Magazine (here’s their awesome tweet from yesterday) will rave about what he is doing efficiency-wise at such a young age. And it’s why, in the end, this “potential” argument is clearly unwinnable. It’s such a fascinating debate and it was a joy to look at and wrestle with what the various stats are saying about Kyrie right now — and possibly at his peak career value. So I hope you enjoy and see you next Wednesday for the next edition of The Diff.


Photo: David Liam Kyle/NBAE via Getty Images

  • The_Real_Shamrock

    Love Kyrie glad he’s here I was wrong when I had Williams over him but Irving is not amongst the top 10 players in the NBA not by a mile. If you want to narrow the field and include age and/or position then that will help but not overall to me.

    I also question whether or not Kyrie can single handedly carry the Cavaliers a la LeBron James. I mean if James couldn’t get it done I have serious doubts whether Irving can. Honestly I’m hoping Irving doesn’t have to do it all himself but right now those prospects aren’t looking good to me. See Jacob’s last stat in the article.

  • mgbode

    Deron Williams – he basically took 2 yrs off and is finding out that he cannot just flip that switch back on.

  • Jack

    Good fun. Wish the silly Kelly soap opera wasn’t getting all the attention. Tough afternoon to debut, Jacob. Still a little confused about usage rate and PER correlation. Isn’t efficiency efficiency? Is PER not designed to measure effectiveness on a per minute basis? Are you just saying that people have the ball more have a better CHANCE to put up a higher PER? I mean what’s Dion’s usage rate? Gotta be fairly high, right? But not a very good PER…

  • Adam Rosen

    Wins, more than defense, will define Kyrie. The front office needs to do something this off season. Seriously!

  • JacobWFNY

    Gracias, Jack. I knew I was getting buried by the Kelly story this afternoon. But I had some more writing to do so didn’t have much of a choice…

    In my opinion — I haven’t formulated this out yet — I believe that usage % and PER are indeed positively correlated. PER is not pure efficiency — the more shots you take, the better you will be. Or at least that’s my interpretation of it, I believe. I think that’s right.

    Example: Two players that are exactly the same in all metrics, but one takes 15 FGA/36 mins and the other takes 6 FGA/36 mins, the one with more shot attempts will have a higher PER. Technically speaking then, PER then doesn’t measure efficiency. It also has the same basic formula as usage %. That’s my theory.

    Dion’s usage % is 25.6. His PER is 12.8. Initially, goodness, that does seem like a peculiar 2/1 relationship. So here’s another B-Ref search: Rookie guards in the last 10 seasons, qualified for MP/G leaderboard, usage % >= 24, PER <= 15 (league average). —–

    The players on this list are fun comparisons: Ben Gordon, Marcus Williams, Nick Young, O.J Mayo, Jonny Flynn, Brandon Jennings, Kemba Walker, Klay Thompson and Waiters. Gordon's not a crazy overall comp. Same with Walker. SGs are usually quite inefficient at PER — lots of long 2's, not many rebounds/assists — so overall, Waiters' low PER isn't that shocking. He's also been much better lately in just about every metric.

  • JacobWFNY

    LeBron’s (career PER of 27.3) ability to carry a team is unprecedented. It’s Magic or Jordan-esque. That’s it. I wouldn’t even put Kobe (only one season with >26.2 PER — ’05-’06) or Durant (only one season >26.2 PER — this year) on that pedestal. Maybe eventually for Durant because of his do-it-all ability, but those skills are typically once-in-a-decade profiles.

    Kyrie’s upper-end potential (i.e. best-case scenario) might be at some point to be the best PG in the league. That would be a top 5 player in the NBA, likely. But that would be for a couple years maybe? Paul is clearly the top PG right now. But does Paul carry a team like LeBron? No, but he’s certainly good enough to lead a stacked team to a great run. I’d expect similar in a best-case scenario for Kyrie.

  • The_Real_Shamrock

    Deja vu I think I heard this one before but actually Williams has been getting back to his more normal game I guess it does matter who his coach is unfortunately.

  • The_Real_Shamrock

    I agree I was simply saying Kyrie wasn’t there that doesn’t mean, as you’ve pointed out, he won’t be eventually, hopefully. As far as CP3 goes I’m bias I love the guy have since his college days and I could argue he leads his team but I won’t have actual stats to back it up. Certainly not to the extent LBJ leads a team which was kinda the whole point of why I commented in the first place. Your opening response to my comment says it all. This is why I wouldn’t mind his returning.