Can Steph carry his own team thread.

The Warriors are about 10 games over a .500 record through 37 games. Adding 10 wins is considered exceptional for a full season, and we’re talking about less than half a season. No, I don’t think there would be 20 players capable of that.

 

LOL NO. First off that is not how win shares work. There is an actual stat for that and curry is currently sitting at 4.8. You claiming he added 10 wins to this team would mean he is having a better season than michael jordan

Second, that obviously just goes to show the warriors supporting cast are FAR better than a .500 team

 

If the Warriors without Curry is a sub .500 team, and with him they are 29-8, then his presence did evidently add 10 wins irrespective of what “win shares” computes to (which is just a box score stat at the end of the day).

Whether that means he is having a “better season than Michael Jordan” depends on how you rate seasons.

 

they arent. They are much better

And no, there is no case for stephen curry being better than jordan

 

Then when you said “that’s not the accomplishment you are making it out to be” you were not actually referring to the case I was talking about in which they would be a sub .500 team. I think they probably would be, but you’re free to disagree and think that 20 other players in place of Curry would have this team at the same record.

 

who were you originally arguing with that agrees the warriors are a sub 500 team? Thats literally just you saying that

 

PER and WS/48 aren't suppose to be used by itself to definitively determine how good a player is compared to another ESPECIALLY when you are using them in small sample sizes such as one playoff run. PER is a decent all-one efficiency aggregate but it has its shares of flaws, in addition to not being very predictive in terms of measuring "wins." https://towardsdatascience.com/problems-with-per-in-the-nba-fa0e611543e2

WS/48 is a useful tool to measure box score production on a rate basis but doesn't account for volume and is not meant to be used by itself to rank individual performance. Neither tool should be used to make definitive judgements especially in small sample sizes. Even BasketballRef which created WS/48 as well as their version of WAR & Box +/- would readily admit that their advance stat are not the most accurate ones publicly available. This is because they are designed only to use box-score data in order to also measure/compare the performance of players in the past (before tracking data). The best ones available to the public currently is epm, apm, raptor, lebron and a few others. Here's how some team execs rated publicly available stats https://hoopshype.com/lists/advanced-stats-nba-real-plus-minus-rapm-win-shares-analytics/

The reason why Curry ranks so highly on many of these all-in one advance tracking stats is due to his production on the court (as well as when is he off the court). The Warriors OFFRTG with Steph's on the court numbers regardless of teammates has been phenomenal since 2013. There was a question of where Steph's influence and Draymonds influence differentiated back from 2014 to16 due to them sharing so many minutes. This led to both having historic RPM and other advance rankings which utilized tracking data. However that question was clarified over the years with the ever increasing minutes they aren't on the court together, especially recently when their substitution patterns has increasingly diverged.

For this season with Curry/Dray on the court the Warriors have a +12.67 Net rating 114.42 OFFRTG in 774 minutes.
With only Steph on/Dray off the numbers are +16.34 Net rating and a 117.63 OFFRTG in 442 minutes.
With only Dray on/Steph off the numbers are -3.18 Net rating & 106.08 OFFRTG in 244 minutes.

This is a very small sample size, so variance, lineup, opponents and other factors can overly influence the results. However this data is consistent with Curry's on court numbers since 2013.

Steph on/Draymond off court in over 5k minutes: Net rating +9.08 Offrtg 115.39
Draymond on/Steph off court in over 5k minutes: Net rating -1.09 Offrtg 104.28
https://www.pbpstats.com/wowy-combos/nba?TeamId=1610612744&Season=2021-22,2020-21,2019-20,2018-19,2017-18,2016-17,2015-16,2014-15,2013-14&SeasonType=Regular+Season&PlayerIds=201939,203110

We see the same occurrence from 2017 to 2019 with Steph and KD.
Steph on/KD off Warriors Net +12.44 OffRtg 117.61 (2k minutes)
KD on/Steph off Warriors Net +2.4 OffRtg 112.61 (2.5K minutes)
https://www.pbpstats.com/wowy-combos/nba?TeamId=1610612744&Season=2018-19,2017-18,2016-17&SeasonType=Regular+Season&PlayerIds=201939,201142

This shouldn't be too surprising since the Warriors offense is built around the strengths of Curry. An offense built around KD strengths would likely yield better numbers for KD. Perhaps even more reflective of Curry's influence is that every teammate that has played at least 500 minutes with Curry has had a noticeable bump in TS% since 2012, except 6. That's 6 out of 41.
https://www.pbpstats.com/on-off/nba...sonType=All&TeamId=1610612744&PlayerId=201939

Curry is still having an enormous impact on the Warriors this year despite a decline in shooting numbers. His RAPTOR, LEBRON, APM & EPM (all use tracking data) are significantly lower than Jokic (who's having an absurd statistical season). But Curry is still leading in total WINS added in some of those same metrics due to minutes played. https://dunksandthrees.com/epm https://www.bball-index.com/2021-22-lebron-data/

Lastly advance metrics are just one tool to evaluate players, none of them should be used (by itself) to evaluate individual players. Ideally you want to use several metrics as well as the "eye test" to make that determination.

 

I made a conditional statement: “And if you don’t believe it to be true, then you think Curry’s presence has turned a borderline playoff team into a top-tier contender.” Note the word “if”.

 

Steph is great, but Chris Paul is still better. Come at me Bros.

 

Edited in bold.

 

Think age is getting the best of everyone. I have never seen Curry shoot this badly in the regular season for this long stretch of games.

He is also shooting a ton more 3's.

GS's system is still pretty incredible though, props to Kerr for putting this around Curry for a whole decade, it sure is hell effective.

 

yes, i undestand that, but they are still good shorthand and the tops of those lists tend to be the very best players (though WS48 is a little wonkier). so when guys like chris paul and david robinson show up there as well, it's not unreasonable to start asking why and if they are simply overrated because a formula puts certain weights on certain things and maybe they are just specially made for those formulas and we should dock them a little, but not necessarily throw out the numbers overall.

my chris paul comment was for career playoffs. i'm pretty sure at one point he was #5 in both for prime years (ages 22 to 35). it was not a small sample for anybody (i think i did 1500 minutes minimum).

PER can't predict wins because the box score can't show defense. but the box score in general over history still shows us the most productive players, especially when at least somewhat accounting for bad shooting and turnovers (people think PER should punish inefficiency even more but it still tends to suss out guys like iverson who could theoretically dominate PER if it didn't punish inefficienty).

like it's not a surprise to me that the top 5 playoff PER's ever (again, ages 22-35, minimum 1500 or 2000 minutes i think) are jordan, lebron, shaq, hakeem, and duncan. depending on how one feels about russell, i think those would be my top 5 as well for the playoffs. it may not correlate to wins, but it seems to correlate to greatness, even with its weaknesses, in a way that on/off numbers don't always seem capable of.

it just feels like those stats have more "are we sure about this?" type moments. like RAPTOR currently says jokic is twice as good as anybody in the nba, van vleet is equal to curry at 4th, and gary payton II, derrick rose, and alex caruso round out the top 8.

on/off RAPTOR has george hill 2nd, immanuel quickley 4th, jarred vanderbilt 7th, and kenrich williams 8th (and none of those are limited minutes for this season, but then we'll get the typical caveat that we need multiple seasons of data and then how much is it worth if we can barely look at a whole year and trust it). it's not just that there are too many of these examples, but the on/off numbers seem to have consistent favorites and non-favorites. like kyle lowry was like 1st, 5th, 6th, and 13th in RPM over the last 4 seasons and i believe pretty good before that. yes, it's good when our assumptions are challenged and we have to think "is lowry actually really great?", but it seems like some people just trend as overrated or underrated in these impact numbers. steph is great but also seems to be in the lowry category of being even more loved than makes sense by the impact numbers.

yes, and so we are to either conclude that draymond green and his DPOY defense and amazing passing aren't really as important as we think and steph is insanely important or that, whatever special circumstance the warriors system and line-up choices present, that they tend to systematically push the value to curry and away from his teammates based on the best ability of on/off metrics to figure out the impact of one person amongst hundred of lineups and opponents line-ups, etc. or maybe it's just as simple as steph being supervaluable in the regular season when the warriors style is impossible to properly prepare for and a decent chunk of that value goes away at playoff time.

and i feel like 2019 is a good example of this, especially as it relates to KD. for years we heard warriors fans tell us that curry was the important guy on the team. after all, the warriors had some crazy record when KD didn't play (like 33-1 or something) and didn't tend to notice KD's absence. then they lose the 2019 finals and it's not fair because KD was hurt and the raptors played a box-and-1 defense and it turns out KD is important after all.

sure, that's what i do. and i guess i will just always assume that the impact metrics that almost always paint curry as the best in the league must be at least slightly off because he has never ended a season being considered the best in the league and therefore i will probably lend at least a decent amount of weight to the other metrics.

 

Whatever Ben Taylor and his ilk say, traditional nba stats used to determine the best players in the league will ALWAYS be the bar.

these stats like on/off and ramp like Kevin Peyton says are too hit and miss. You said it yourself Gary Payton ii, Otto porter, quickley are seen as the best players in the league…. Good luck convincing someone of that

All it is, is an attempt by Ben Taylor to find stats that uplift his boy toy steph curry. Nobody on Reddit except the curry fans take that guy serious anymore

 

As players age, they usually drive less and less. It's natural that you are gradually slowing down and driving needs speed.

 

These threads will always exist. He carries the team as much as the team carries him.

It is the perfect synergy and he isn’t built or wired like Harden to put up insane stats most of the time but he complements fellow Allstars, good players as well as they complement him.

Intangibles that you cannot measure precisely.

 

The reason PER list of greatest players/performances feels "right" is because Hollinger tailored his stat to match Jordan to other historic performancers. In other words, he cooked the numbers. Despite this I like PER as a shorthand also, but it isn't a very good advance stat compared to what's available. There are numerous articles by data scientists and analytists (much more verse on the subject than me) who will attest to that.

Raptor, Lebron and other advance stats that combine box score + tracking data are going to have some wonky results, ESPECIALLY in the on/off or offense/defense columns as the data is more narrowed. Looking at their total wins added would probably give a more accurate measure. But like any advance stat, you're going to have some outliers or aberrations. Furthermore their on/off numbers perhaps don't do a good enough job of accounting for lineup and roster differences.

As for RPM and tracking stats needing multiple seasons required for accuracy. RPM was very well received when it was first started in ESPN, but less so after the creators left and ESPN continually changed their formula around 2018. One of its original creators is a big fan of EPM.
As for why tracking data is better with multiple seasons, its primarily due to measuring defense which is by far the hardest to capture. the larger the sample size the more accurate the results. This however doesn't mean one season's data from epm or apm or raptor etc isn't going to be useful or accurate, especially as it pertains to measuring a players impact on the score when they are on the court. If you are interested their are many articles written which compares and evaluate how accurately these metrics translate to wins.

These raw pbp and on/off data are not pushing or choosing one player or another, it's simply raw box score data showing what happens when they are on or off the court. If you look at Draymond's pbp data, he has consistently posted elite defensive numbers when he is on the court. You don't need some advance algorithym or regression to see that. Unfortunately for Draymond, the team's offense is terrible when he is on the court without Curry. It is pretty easy to see the impact of individual players simply by using raw pbp data without any advance formulas. WIth enough minutes you can get a fairly accurate measure of a players impact by tracking their stats with/without other variables (players). The advance stats like apm, epm, raptor etc just use that raw data and put it into the algorithyms and regression analysis, along with adding box score stats to create a more refined tool.

As for KD and the Warriors and regular season vs post. As I stated in an earlier post. The Warriors system is built around Curry and maximizes his strengths. It's not surprising that KD joining such a system would be less impactful than Curry during the course of the regular season. Systems is just as important as talent in how impactful a star will be (especially unique ones). Nash in the Suns system was a perennial MVP. In the Dallas he was merely an All-Star. Harden in MDA's Rockets was an MVP and posted some of the highest scoring volume and usg in league history while in Mchale or KD nets is much less so.

The Playoffs and the regular season are two different animals. People really shouldn't use regular season stats or advance metrics from both interchangeably. They aren't the same thing. Most players/system regress in the playoffs due to the increase intensity, effort and game planning. Curry and KD's impact numbers suddenly became a lot closer, because the Warrior's system was no longer as effective. KD's individual sot creation became a lot more valuable. It shouldn't be a surprise that the 2019 Warriors certainly needed KD or atleast Klay in order to beat the Kawhi Raptors.

With the exception of the 2016 season and that moment before the end of game 7, Lebron has been the considered the best player in the World since 2010 or 2011.

As for Curry being rated as the best constantly by impact metrics. I'm not sure that is correct, atleast not in the past few seasons. Raptor, Lebron, APM, EPM all have Jokic as the best player the last two seasons. Curry is usually amongst the best because of how he affects his teams offense, such as this incredible raw statistic. Eevery player who's shared the court with Curry for over 600 minutes has had a postive increase in TS. Every player who's played over 1500 minutes with Curry has seen their TS increase by atleast 5%.

 

Oh, I was under the impression that people were debating how much better Curry makes the Warriors.