Curry 22' vs Harden 18'

[durantbird]

Who's the better player? This year's Steph or MVP Harden?

[jasonxxx102]

I will take every year of Steph’s entire career over MVP Harden.

If you just lay the stats side by side with no names and no context it’s absolutely Harden but Curry is just more important to winning basketball

[f4p]

I will take every year of Steph’s entire career over MVP Harden.

If you just lay the stats side by side with no names and no context it’s absolutely Harden but Curry is just more important to winning basketball

Well 2018 Curry seemed less important to winning basketball when he was playing with 3 other prime Hall of Famers and was still down 3-2 to 2018 Harden with slightly post-prime Chris Paul. I would think the better player with the much better teammates wouldn't need a lucky break to come out on top in a playoff series.

[Max123]

I will take every year of Steph’s entire career over MVP Harden.

If you just lay the stats side by side with no names and no context it’s absolutely Harden but Curry is just more important to winning basketball

What do you mean by Curry being more important to winning basketball? His teams surely have won more than Harden's but besides a single individual's impact there are lots of other factors that result in team performance. How have you come to this conclusion?

[Doctor MJ]

Who's the better player? This year's Steph or MVP Harden?

So, one of the reasons I struggle with participating in thread where they ask "The best X seasons between Player A and Player B", is that when we talk about guys who have different approaches, the question that's more meaningful to me is who has the better approach for winning titles.

In general, I think Curry's approach scales a lot better for winning titles than Harden, so I'm inclined to pick Curry in basically any season where Curry is doing his thing.

[Cavsfansince84]

This thread/question to me strikes at a fundamental issue I see with comparing players and how we view rs as compared to ps. Because I think we could all agree that Steph just had his worst rs since about 2013 but since he also just won a title while having a very strong playoffs his rs of like 65 games just gets conflated with his playoff run of like 22 games. While Harden in 2018 without a doubt had a stronger rs but lost in the wcf in 7 games so I think people are just going to naturally side with Steph here. While if we just compare playoff runs people would be like 'why is this even a question' yet my point is that's fundamentally what we are doing anyhow.

[Doctor MJ]

This thread/question to me strikes at a fundamental issue I see with comparing players and how we view rs as compared to ps. Because I think we could all agree that Steph just had his worst rs since about 2013 but since he also just won a title while having a very strong playoffs his rs of like 65 games just gets conflated with his playoff run of like 22 games. While Harden in 2018 without a doubt had a stronger rs but lost in the wcf in 7 games so I think people are just going to naturally side with Steph here. While if we just compare playoff runs people would be like 'why is this even a question' yet my point is that's fundamentally what we are doing anyhow.

You raise a good point, but do consider this:

Here's the raw +/- of the Rockets in the regular season in '17-18:

Eric Gordon +585
PJ Tucker +526
James Harden +520
Chris Paul +499

And for the Warriors this season:

Steph Curry +507
Jordan Poole +324
Andrew Wiggins +265
Otto Porter +232

It's not hyperbole to argue that Curry even in his regular season form this year was more impactful than Harden in '17-18.

[Cavsfansince84]

You raise a good point, but do consider this:

Here's the raw +/- of the Rockets in the regular season in '17-18:

Eric Gordon +585
PJ Tucker +526
James Harden +520
Chris Paul +499

And for the Warriors this season:

Steph Curry +507
Jordan Poole +324
Andrew Wiggins +265
Otto Porter +232

It's not hyperbole to argue that Curry even in his regular season form this year was more impactful than Harden in '17-18.

Let me ask you this, are people so infatuated with +/- solely based on its usefulness/merit or because it tells them what they want to hear about players? That's not so much an accusation against you personally as it is a question I think needs to be asked because its human nature to use data(or in bb metrics) which reaffirm what we want to think is true. That's part of why I hate using just 2 or 3 metrics much less 1.

[Ginoboleee]

Who's the better player? This year's Steph or MVP Harden?

So, one of the reasons I struggle with participating in thread where they ask "The best X seasons between Player A and Player B", is that when we talk about guys who have different approaches, the question that's more meaningful to me is who has the better approach for winning titles.

In general, I think Curry's approach scales a lot better for winning titles than Harden, so I'm inclined to pick Curry in basically any season where Curry is doing his thing.

Precisely.

My personal approach would simplify/fudge this away from Approach/Style (for which Dr. MJ is right, yet again!) towards these 2 questions:
1. Is the championship-winning player in question on the 1st or 2nd team All-NBA? And thus clearly a top 10 player, at least by voter acclaim.
2. Did the regular season MVP in question have a disappointing playoff run? Any particularly rough elimination games?

If the answer to both questions is yes, even in a cross-time analysis, then the conclusion is as follows:

3. The ring-winning top 10 player had a better season than the playoff-disappointing RS MVP. Because this is a team sport, even though we continue to implicitly pretend that it is not.

Now, regarding the subtle disconnect between (a) who had a better season, and (b) who is actually a better player, well that is a larger question that requires even more transparency for standards/evaluations, and further nuanced discussions.

[Doctor MJ]

You raise a good point, but do consider this:

Here's the raw +/- of the Rockets in the regular season in '17-18:

Eric Gordon +585
PJ Tucker +526
James Harden +520
Chris Paul +499

And for the Warriors this season:

Steph Curry +507
Jordan Poole +324
Andrew Wiggins +265
Otto Porter +232

It's not hyperbole to argue that Curry even in his regular season form this year was more impactful than Harden in '17-18.

Let me ask you this, are people so infatuated with +/- solely based on its usefulness/merit or because it tells them what they want to hear about players? That's not so much an accusation against you personally as it is a question I think needs to be asked because its human nature to use data(or in bb metrics) which reaffirm what we want to think is true. That's part of why I hate using just 2 or 3 metrics much less 1.

I'm never telling people to only use one metric. I am asking you to explain to yourself why Curry shows such separation in a metric while Harden does not.

I'll tell you, lately I've been getting people getting annoyed at how often I use raw +/- to start conversations with folks. They often interpret me using a particular stat as if I'm using one stat to do my analysis (and are then confused when my final assessments don't match what any one particular stat appears to say).

The question I find myself asking is: How should I be trying to get people thinking about stuff?

If I post everything I've thought about in any of these topics, we're talking about posting an encyclopedia. (That's not me saying I'm so special here, many of us could say the same.) Posting too many details generally leads people to shut down. Even if they end up typing "Okay, you're right.", it's not making them think through all that needs to be thought through so much as just deferring (at least publicly) to someone they don't want to challenge.

So how do I jump in and make people think about things they haven't thought about?

The answer is never going to be based on stats like PPG which everyone already knows.

Similarly, using extremely abstract stats that purport to "factor everything in" isn't going to do much.

Using raw +/- gives me something that a) everyone understands on a fundamental level but b) they generally haven't actually thought through.

In your case, likely either this stat is a surprise to you, or you've seen it before and have chosen not to think it through (possibly because it disagrees with conclusions you've already come to). In either case, I'm challenging you to think through what happened beyond simply "noise". Noise is certainly a thing to be considering, but if it's your only explanation, then you're learning nothing about basketball in your analysis, and if it's your default assumption, then you're actively preventing yourself from learning more about basketball (along this avenue) when you analyze basketball.

Also, there's a qualitative analogue to this with stuff like "gravity" and "latent value". I can't tell you how many people I've seen talk about gravity as if it's something that doesn't actually exist, which is amazing because I can demonstrate it to someone who knows nothing about basketball in minutes.

I think in general people tend to put walls up to protect their confidence in their approach, and so when something comes along where it's hard to know how to factor it in with other things, they find reasons why they don't have to - both consciously and subconsciously. It's a thing the brain does as it struggles to make sense of the world and not something that should be seen as horribly damning...but it's something to be watching out for.

[Dr Positivity]

Curry is simply a better player than Harden. You'd have to go back to 14 before I'd consider taking prime Harden over him.

[No-more-rings]

It’s totally predictable for most to pick Curry based on his playoff run, and fresh off it at that. I’d also suggest that people not brush Harden aside so easily. Some things to consider:

- Harden was the best player on a 65 win and 8.21 SRS team
- The Rockets had a better ORTG than Curry’s Warriors did
- Harden was arguably a Cp3 injury away from beating Curry’s team who had a healthy KD, Klay, and Draymond behind him
- Despite the offensive dropoff in the playoffs, Harden played some of the best defense in his career which helped contribute to defense that stiffled the Curry/KD Warriors like we’ve never seen before

I’d like to know if Curry is so easily better or whatever the case, i’d like to know why that younger Curry’s Warriors didn’t win that series more comfortably. I don’t know how you’d make the case that Curry’s cast was somehow worse than Harden’s. Harden also easily outplayed him the following year, but again lost because Curry had the superior cast. Curry only had a 54 ts% in that series, but similar to a lot of series in his career he gets bailed out.

[Cavsfansince84]

I'm never telling people to only use one metric. I am asking you to explain to yourself why Curry shows such separation in a metric while Harden does not.

I'm not denying that there is validity to +/- or that it maybe taps into impact which goes beyond other numbers. In relation to Harden's +/- in 2018 I think Steph got plugged into more different lineups than Harden did most likely and that just shows that Steph is a much better player than they are and was the engine of his team. Does this mean Steph was better in 2022 than Harden was in 2018? I'm not sure that's a correct way to interpret that. We do know that when Harden and CP3 both played they were a ridiculous 51-4 or something like that which speaks to me of huge impact as well. We also know that there was a huge stretch of games where Steph was bad and GS was losing. I take that into account also regardless of how the other numbers may paint the season on a whole.

[f4p]

It's not hyperbole to argue that Curry even in his regular season form this year was more impactful than Harden in '17-18.

Yes, it is. So now we're not even falling back on the ol' "Harden sucks in the playoffs" chestnut (when Curry falls off just as much), but now he doesn't even really have a lot of impact in the regular season? Compared to one of Steph's worst seasons? I mean going by plus/minus, I guess Harden wasn't quite as good as PJ Tucker and Eric Gordon. Steph being great can't just be completely independent of his own level of play. Steph plays like 2016, he's amazing and better than prime Harden. Steph has his worst season since his rookie year, he's still apparently amazing and better than prime Harden. If his plus/minus doesn't even care if he makes his shots, apparently he should just stop trying and focus on gravity.

Harden led a 65 win team that, when healthy, is a top 15-20 team in NBA history and was practically unbeatable (42-3 when Harden/CP3/Capela played in the regular season), and that was beating Steph + KD + future hall of famer Klay + future hall of famer and reigning DPOY Draymond with CP3 putting up 20/6 on 52 TS%. you consistently see KD and Steph called way better in the playoffs than Harden. That would mean the Warriors not only had the 2 best players in the series, already almost a surefire win, but comfortably the 2 best players in the series, with 2 more hall of fame teammates. That should be guaranteed sweep or gentleman's sweep territory, like Lebron and Kyrie running over the East type stuff. Not down 3-2 and being saved by an injury territory. Unless the argument is that 2018 Curry was not that good and 2022 Curry is so much better that he's above prime Harden, it doesn't make sense (and you basically said that wasn't the argument by saying Curry is always better).

[Doctor MJ]

I'm never telling people to only use one metric. I am asking you to explain to yourself why Curry shows such separation in a metric while Harden does not.

I'm not denying that there is validity to +/- or that it maybe taps into impact which goes beyond other numbers. In relation to Harden's +/- in 2018 I think Steph got plugged into more different lineups than Harden did most likely and that just shows that Steph is a much better player than they are and was the engine of his team. Does this mean Steph was better in 2022 than Harden was in 2018? I'm not sure that's a correct way to interpret that. We do know that when Harden and CP3 both played they were a ridiculous 51-4 or something like that which speaks to me of huge impact as well. We also know that there was a huge stretch of games where Steph was bad and GS was losing. I take that into account also regardless of how the other numbers may paint the season on a whole.

You've good thoughts here, and definitely the "whose better argument" is always a trickier one than the "who was more valuable" argument, but I note that you're not addressing the fact that basically GS with Curry over the course of the season was about as effective as Houston with Harden.

Now, you can talk about stuff like Houston having the #1 seed and not having to max themselves out in the regular season, and I don't want to dismiss this aspect of things...but we shouldn't ignore the fact that having a +/- in the 500-600 range is nowhere the cap of what we've seen from players in the past.

If we go look up raw +/- numbers for all the seasons on bkref (back to '96-97), you won't find Harden in the Top 100 at all.

It needs to be reconciled, when we look to understand how good the Harden-led Rockets were, that we don't see outlier +/- numbers from Harden like we'd expect from his W/L record.

And of course, all of this would matter less if Harden were known for coasting in the regular season and turning it on in the playoffs, but it's quite the opposite.

[Doctor MJ]

It's not hyperbole to argue that Curry even in his regular season form this year was more impactful than Harden in '17-18.

Yes, it is.

Okay, I stand corrected. Thank you for enlightening me. Now that I understand basketball so much better, I'll leave the thread.

[f4p]

This thread/question to me strikes at a fundamental issue I see with comparing players and how we view rs as compared to ps. Because I think we could all agree that Steph just had his worst rs since about 2013 but since he also just won a title while having a very strong playoffs his rs of like 65 games just gets conflated with his playoff run of like 22 games. While Harden in 2018 without a doubt had a stronger rs but lost in the wcf in 7 games so I think people are just going to naturally side with Steph here. While if we just compare playoff runs people would be like 'why is this even a question' yet my point is that's fundamentally what we are doing anyhow.

It's not even like Steph was so amazing in these playoffs compared to his other years that are sometimes viewed as not that great. The Warriors once again had the #1 defense in the league (likely understated from all he games Draymond missed) just like in 2015 (#1), 2017 (#2), and 2018 (slacked in the regular season but #1 playoff defense when they tried). They were more talented than all of their opponents, paying an enormous luxury tax bill to give someone like Wiggins $25M to be their 4th or 5th best player (if he's your best player, you're in trouble; if money is no option, he's great to have around for perimeter D and he can just drift in and out on offense like he always does), and having their 6th man Poole be so good that he literally averaged 28.7 ppg on 84 TS% in his 3 playoff starts, a 3 game stretch that is likely among the best in NBA playoff history for volume/efficiency.

With all of that, and with the team finally healthy by the time the playoffs started, Steph didn't have to do anything out of this world to win a title. Certainly it didn't take some 94 Hakeem/03 Duncan/19 Kawhi type performance. Basically show up with the usual 27/6/5, 24 PER, 59-61 TS% playoff performance from the non-2017 portion of his playoff career and there wouldn't be much of a struggle. And there wasn't. Even when the Finals got iffy, they just followed the 2015 script and smothered the other team for 3 straight games, even winning the game where Steph broke his incredible playoff 3 point streak.

Is Steph's 24.4 PER, 0.203 WS48, +6.5 on/off per 100 (if we're going with plus/minus) any better than Harden's 24.9 PER, 0.163 WS48, +20.1 on/off per 100, especially considering Harden spent 12 of his 17 games against the #1 regular season defense (utah) or #1 playoff defense (warriors)? Is beating the 4 teams Steph beat a bigger accomplishment than going up 3-2 on arguably THE superteam, only to have the rug pulled out from under you?

[tsherkin]

Posting too many details generally leads people to shut down. Even if they end up typing "Okay, you're right.", it's not making them think through all that needs to be thought through so much as just deferring (at least publicly) to someone they don't want to challenge.

Ah, THAT'S what I'm doing wrong.... xD

[Cavsfansince84]

You've good thoughts here, and definitely the "whose better argument" is always a trickier one than the "who was more valuable" argument, but I note that you're not addressing the fact that basically GS with Curry over the course of the season was about as effective as Houston with Harden.

Now, you can talk about stuff like Houston having the #1 seed and not having to max themselves out in the regular season, and I don't want to dismiss this aspect of things...but we shouldn't ignore the fact that having a +/- in the 500-600 range is nowhere the cap of what we've seen from players in the past.

If we go look up raw +/- numbers for all the seasons on bkref (back to '96-97), you won't find Harden in the Top 100 at all.

It needs to be reconciled, when we look to understand how good the Harden-led Rockets were, that we don't see outlier +/- numbers from Harden like we'd expect from his W/L record.

And of course, all of this would matter less if Harden were known for coasting in the regular season and turning it on in the playoffs, but it's quite the opposite.

I see a lot of holes in +/- that just seems innate to the stat that I hate to even go through it all. Like I said, it has value but as we all know there is noise so if I'm looking for some kind of truth through this one data set I think it's already a flawed approach. You did say above that you don't ever just use one metric so my question to you is is there any other metric out there which would put Curry on Harden's level if not above for these two seasons? Otherwise it would seem you are very much relying on just one metric here.

[f4p]

If we go look up raw +/- numbers for all the seasons on bkref (back to '96-97), you won't find Harden in the Top 100 at all.

It needs to be reconciled, when we look to understand how good the Harden-led Rockets were, that we don't see outlier +/- numbers from Harden like we'd expect from his W/L record.

So what's the alternative, that Harden isn't actually that impactful, even at his very best, even in the regular season because his plus/minus isn't all-time?

The 2015 Rockets win 56 games in a western conference with 67, 55, 55, and 55 win teams without half a season from Dwight Howard. With Montiejunas, Beverley, and 37 year old Jason Terry being 2/3/4 on the team in minutes? Is that somehow without Harden having a huge impact?

The 2017 Rockets win 55 games with Eric Gordon and Ryan Anderson as their 2nd and 3rd best players, fresh off a season where Gordon/Anderson were the 2nd and 3rd best players on a team that won 30 games, even with Anthony Davis being the best player (and before someone says injuries, they were slightly worse than their overall record when all 3 played).

The 2018 Rockets surprise with an amazing 65 win regular season where they almost can't be beat when healthy, but that's everyone else?

The 2019 Rockets take off when Harden starts averaging 40 ppg and almost gets an injured squad to the 2 seed, but that's not Harden?

The 2020 Rockets are the last team to look good with Russell Westbrook playing big minutes.

The 2021 Nets go 19-17 without Harden but an incredible 29-7 with Harden, despite only 8 of those games actually featuring the Big 3. And for good measure, the 2021 Rockets have one of the worst finishes to a season in NBA history.

Even the horrible version of Harden in 2022 still made the 76ers look better once he got there and the 2022 Nets imploded in spectacular fashion after he left, even with the vaccine mandate for Kyrie being dropped.

There's simply too much regular season impact over too many seasons and too many different rosters for me to think nearly the peak version of Harden on one of the better regular season teams ever wasn't actually as good as nearly the worst version of Steph Curry. You'd have to have Curry like 4th all time or something before I would believe that about a player.