Peaks project update: #12

Moderators: Doctor MJ, trex_8063, penbeast0, PaulieWal, Clyde Frazier

DatAsh
Senior
Posts: 627
And1: 356
Joined: Sep 25, 2015

Re: Peaks project update: #12 

Post#81 » by DatAsh » Wed Aug 7, 2019 4:52 am

Clyde Frazier wrote:Ballot #1 - 76 Dr. J
Ballot #2 - 64 Oscar
Ballot #3 - 16 Curry

--------------------

Ballot #1 - 76 Dr. J



I get it. It's a 5 minute clip, but I still think you can tell just how talented this guy was that year. An unstoppable offensive force leading his team to the championship. Nets also ranked 1st in defense that season.

For those who doubt the ABA, check out his per 100 #s in 76 vs. 80:

http://www.basketball-reference.com/players/e/ervinju01.html#per_poss::none

They’re nearly identical including efficiency. This is when he was given a bigger role in the offense after Cunningham came aboard as coach.

It’s possible his ball handling is being underrated here due aesthetics. He kinda slapped the ball down as he dribbled, especially on the fast break. Similar to the way Barkley dribbled in his Sixers days. While it may have looked a little sloppy, I think it was just as effective given his big hands and long strides once he went to make his moves.

Also, his ability to get off shots at the rim in tight spaces was pretty incredible. This also had a lot to do with his body control.

The below footage is from 74, but it's pretty similar to the way he was playing in 76.



Ballot #2 - 64 Oscar

Oscar's 64 season was very impressive on a number of levels:

RS: 31.4 PPG, 9.9 RPG, 11 APG, 48.3% FG, 85.3% FT (league leading on 11.9 FTAs per game), 57.6% TS (+9.1% vs. league avg), .278 WS/48

PS: 29.3 PPG, 8.9 RPG, 8.4 APG, 45.5% FG, 85.8% FT (12.7 FTAs per game), 56.8% TS, .245 WS/48

The royals ranked 2nd in SRS that season, losing in the playoffs to the #1 ranked SRS and eventual champion celtics. While his raw averages can certainly be attributed to the fast paced play during that era, his overall efficiency and ability to get to the line at will is pretty staggering.

Oscar's playoff #s do drop slightly across the board, but there's nothing there to suggest that he struggled. His best teammate Jerry Lucas had a serious drop off in scoring and efficiency come playoff time (17.7 PPG on 57.8% TS in RS vs. 12.2 PPG on 43.8% TS in PS). That very well could've been the difference in the series.

63-64 was his 4th season, so the below footage should be able to capture his style of play at the time:



[Yeah... I could do without the music]

What stands out to me is his precision when he makes his moves as well as his strength when he gets inside. Reminds me of west, too, although he wasn't quite as powerful.

Oscar would win also win MVP that season in dominating fashion. Via NY Times:

Oscar Robertson, the Cincin­nati Royals' talented back‐court man, yesterday was voted the President's Trophy, the Na­tional Basketball Association's most valuable player award, by the biggest margin on record.

The voting is by N.B.A. play­ers, with the restriction that they cannot vote for members of their own teams. Robertson received 60 of a possible 85 first‐place votes. In the point scoring on a 5, 3, 1 basis, Robertson received a total of 362 points, a record.

Wilt Chamberlain of San Francisco, who won the trophy as a rookie in 1960, placed see­ond in the voting with 19 first­place votes and 215 points. Bill Russell of Boston, the winner for the last three years, was third with 11 firsts and 167 points.


64 slightly edges out 63 to me overall due to a better individual regular season and team performance.

Ballot #3 - 16 Curry

Arguably the greatest offensive regular season ever, or at least in the modern era, say since 1980? The finals "collapse" doesn't kill the season for me. It literally came down to the last minute to decide the championship. This wasn't a 4-1 trouncing or something ala pistons lakers in 04. I have to take a closer look at 2017 because it's certainly close, but that 2016 season was a sight to see. Appointment television every night and curry was the center of it. It was special.


Taking into account the era, Oscar's TS% is incredible.
FrogBros4Life
Sophomore
Posts: 138
And1: 155
Joined: Dec 30, 2018

Re: Peaks project update: #12 

Post#82 » by FrogBros4Life » Wed Aug 7, 2019 5:10 am

penbeast0 wrote:
FrogBros4Life wrote:
euroleague wrote:...
As for the points disparity in 94....yes, on the surface it looks like Robinson was A LOT more efficient. But again....free throws. Robinson took almost 12 free throws a game that season compared to 7 for Ewing. He's taking 1.5 more FGA and FIVE extra free throws a game. Robinson was only 1 percentage point better in raw FG% that season....


Not sure what you are trying to get at with this statement.

At the risk of being Captain Obvious here, free throws are a part of efficiency. Efficiency tries to measure how much scoring is generated per shot. Obviously drawing FTs and making them is one of the key determinants in how efficient you are scoring, just like adding more weight to a made 3 pointer than a made 2 pointer. TS% adds a less than perfect extra step to try to account for the fact that the NBA doesn't count a missed shot when you are fouled in your shot attempts but it's clearly a lot more accurate measure of points scored per shot attempt than raw FG%.



Well at the risk of also being Captain Obvious here...

1.) Because the statement "Robinson scored 5ppg more from +1.5 FGA" is not as revealing as the more informative "Robinson scored 5ppg more from +1.5FGA and ~+5FTA".

Obviously it's more efficient for the team to be awarded free shot attempts while the clock is stopped so long as the player shooting the free throws is at or above a certain percentage (relative to the team's average points per possession), but because free throws are both circumstantial, and at the sole discretion of the referee, I would stop short of saying players who make a lot of free throws are ---> more <--- efficient scorers, in a general sense. (I can make an argument that prime Shaq was a more efficient scorer than 09 Jose Calderon despite TS% saying otherwise).

I mean, it's probably true in more cases than it isn't, but not enough for me to be comfortable throwing something like that out as a blanket statement. For example, we don't know if all of Player X's (let's use Robinson in this case) free throw attempts were coming on actual shot attempts as you're claiming. What if the opposing team was in the penalty and Robinson was awarded 2 free throws because a referee called a foul on his defender while jockeying for position before Robinson even touches the ball? What if Robinson gets this whistle 2-3 times per game? I don't know if someone scoring 3-5 points per game out of scenarios where they never even touched the ball, much less attempted a shot within the flow of the offense, is an accurate way to reflect someone's overall scoring efficiency. I think you can somewhat separate being an efficient FT shooter from an overall efficient scorer here, if that makes sense. In that case, at least some of the "efficiency" there is credited to the referee for putting Robinson in a position to score points for his team when he otherwise was not necessarily in a position to score period (efficiently or inefficiently). What if Robinson is blocked cleanly at the rim but the referees give him the foul call anyway and he makes 2 FT's? I wouldn't say getting your shot blocked, and erroneously being awarded made free throws as a result adds anything to a player's true efficiency. That's just the referee making a bad call.

And things like that happen with all players, not just Robinson, or Player X. Now, if they were shooting more or less the same amount of free throws this year, I wouldn't think it would be as big of a deal. But 5 free throws a game more is quite a large difference to the point where I think at least some of that disparity is due to factors like the ones listed above. This is why I wade carefully in the waters of statistics without any other context.

and 2.) The statement I was responding to was literally "Ewing scored about the same on a per100 basis but with far less efficiency. Basically, the statement "Robinson scored 5 more ppg on 1.5 more FGA" implies a much greater degree of efficiency on Robinson's part, even if it was not directly preceded by that actual statement. But the actual statement was made. And this is only true if we ignore the FT disparity. I understand the difference between FG%, eFG% and TS%. In 94 Ewing was actually slightly more efficient from the free throw line than Robinson (about 77% to 75%). But because Robinson took ~5 extra FT's a game, this outweighs the slight absolute efficiency Ewing had from the stripe.

The statement "Robinson scored with far better efficiency than Ewing" can be proven untrue by any shooting % metric.
FG%: Ewing = 49.6 Robinson = 50.7 (Robinson +1.1)
2pt%: Ewing = 49.8 Robinson = 51 (Robinson +1.2)
eFG%: Ewing = 49.7 Robinson = 51 (Robinson +1.3)
TS%: Ewing = 55.1 Robinson = 57.7 (Robinson +2.6)

So even using the least judicious of these metrics given the context, Robinson still only ranks out as 2.6% better. I would say that is better efficiency, certainly. But I would hardly say it was "far less efficient" on Ewing's part, as was directly stated. The TS% here is also slightly skewed due to Robinson's career high 35% shooting from the 3 point line this year, on a career high # of attempts (.4), which is still quite minuscule in actual produced value (Robinson added 30 points over the course of the year to his total points scored for the season due to 3 pointers). Meanwhile, Ewing shot 29% from 3 on .2 attempts per game (mostly scramble plays and end of shot clock heaves), leading to 12 points on the year from made 3's. That's an 18 point difference in Robinson's favor, which over the course of the season metes out to around .23 more points a game from made 3's. A difference of about 1/4 of one point per game doesn't really move the needle for practical purposes, but it does add enough weight to make Robinson's TS% a little more superficially pretty than the other shooting % metrics might suggest, especially when combined with his massive lead in FTr (57% to 39%).

So in 94 Robinson scored 5.3 more points per game than Ewing (non pace adjusted), while making 3.1 more free throws a game. That leaves a margin of 2.2 points per game (about 1 basket -- which is basically what happened. Ewing made 9.4 FGs a game, and Robinson made 10.5 -- a difference of about: 1 basket) that I believe can truly be ascribed to Robinson's slight, but clear efficiency advantage if we are looking at the context honestly. And that looks to be about right by judging the actual % numbers as well. Robinson scoring +5ppg on +1.5 FGA and the same # of FTA would be "much better efficiency", assuming they still shot free throws at about the same %. But +5ppg on +1.5 FGA AND +5 FTA isn't anywhere near as efficient comparatively speaking. If it were, there would be a much larger discrepancy in their shooting % numbers than actually exists. So the statement, "Robinson scored much more efficiently" just isn't backed up by any data. 2.6% at face value isn't "much more efficient", even before we account for any artificial inflation. More efficient? Absolutely. But "much" more is a bit of a stretch.
FrogBros4Life
Sophomore
Posts: 138
And1: 155
Joined: Dec 30, 2018

Re: Peaks project update: #12 

Post#83 » by FrogBros4Life » Wed Aug 7, 2019 5:29 am

trex_8063 wrote:1st ballot - '95 David Robinson
The near-reality as I see is that David Robinson was asked [by the Spurs] to be Bill Russell on defense and simultaneously be Shaquille O'Neal on offense.......and he kinda takes some flack for not being up to the task [primarily in the playoffs]. But realistically, if he'd been consistently capable of maintaining his rs standard of offensive performance and efficiency during the playoffs, we'd have been discussing him in the top 3 positions of this project. So I don't think it's off base to give him some consideration now around #10. This version of Robinson anchored a -2.9 rDRTG (5th/27) with a principle cast [in descending order of minutes played] of Avery Johnson (scrappy and energetic, but seriously undersized even for a PG; mediocre defender overall), Sean Elliott (mediocre defender), Vinny Del Negro (probably slightly weak defensively, iirc), Chuck Person (a pinch past prime, never a good defender anyway), Dennis Rodman (erratic defensively [awful in the Houston series, fwiw], and missed 33 games), and JR Reid, Terry Cummings, post-prime Doc Rivers (Rivers probably the only one of those three I'd say was passable good defensively [edit: +/- maybe Cummings]).
This version of Robinson simultaneously anchored a +3.4 rORTG (5th/27) with the aforementioned cast; they won 62 games (+5.90 SRS) overall. Made it to the WCF where Dennis Rodman had a total [and very public] meltdown, and the Spurs lost the series to Houston (with Hakeem in God-mode) in six games (outscored by a grand total of 10 pts in the entire series). Typically stated as Hakeem owning DRob and making him a helpless play-thing, though it's rarely acknowledged that Hakeem [because of how their offense and roster was structured] largely enjoyed single coverage (by Robinson), while Robinson was largely guarded by Olajuwon + 1-2 friends.
It's rarely acknowledged that DRob's cast [which had shot 37.5% from beyond the arc in the rs] somewhat crapped the bed shooting just 31.9% in this series (and did I mention they were only outscored by 10 points total in the entire series?); and again Rodman's meltdown and poor play is rarely given light of day in the construction of the usual narrative.
jsia, I think he deserves a look around now.





1.) You say that David Robinson was asked by the Spurs to be both Bill Russell on defense and Shaq on offense. I think that's a true statement. Do you also feel that is true of both Ewing and Olajuwon or do you think Robinson's responsibilities were somehow more grueling? You also say you think Robinson receives some undue flack for this (I can also agree with this). Do you think Ewing's flack for that same situation is equally undue? more? less? Or do you feel in his case it's flack that has merit?

2.) I just want to say that even though I made a post yesterday that ran counter to some of these points, I still think your overall argument here is totally valid (on Robinson vs Hakeem, on Robinson 95, on Robinson's career etc.). It's interesting how what appear to be polar opposite stances, realistically probably have at least one foot each in the realm of truthfulness.
liamliam1234
Senior
Posts: 679
And1: 663
Joined: Jul 24, 2019

Re: Peaks project update: #12 

Post#84 » by liamliam1234 » Wed Aug 7, 2019 5:39 am

It is really funny, EB, how you are still going on with this. Again, what does it say if your ego is so fragile that you refuse to acknowledge anyone not immediately accepting every little point you make?

E-Balla wrote:If he wanted to gather information basketball-reference is right there and he already knows the criteria being used. I'm not one for the gish gallop.


Yes, totally a gish gallop for me to not want to redo your work from scratch. I know whenever I read any academic study or data analysis, the authors always say, “Hey, if you want to look at some extra variables of our data set, screw off, you are on your own.”

He clearly didn't read my posts, why gather more statistics for him to write off and ignore?


Telling that this is all you can ever say. Maybe we should all follow in your lead. “Oh, you called my brilliance into question? You just must not be reading properly.” And it is especially convenient when you using it as an excuse to willfully misread and misinterpret every post against you. Tell me, does this perpetual sense of victimisation keep you up at night?

E-balla wrote:
Blatantly dismissing metrics because of his “eyes do a better job” lol.

"Metrics". I'll come up with a number called super player rating made by averaging their place on each NBA leaderboard. That's now "metrics" worth discussing no matter how worthless?


Probably would do a better job than the vast majority of eye tests. But as you have incessantly sought to establish, your eye test must naturally be flawless.

E-balla wrote:
I’m not sure if requesting to know the source of the data qualifies as shifting burden of proof.

Find one post prior to this where anyone mentioned needing a source?


Totally a good-faith interpretation of his point. Yes, we all know you “sourced” the information from basketball-reference. Maybe contrast your posts with Frog’s and see if you can figure out the difference. Oh, wait, you are so much of a narcissist you literally cannot.

E-balla wrote:
What’s to stop any/everyone from making up random stats and then forcing the other party to do the math to refute it? Liamliam1234 requested sources and wanted to know specifically which teams/series the stats applied to.

You did a fantastic job of actually presenting further context which does confirm E-balla’s original claim about their peak playoff performances.

Nah he has a patience for bull.

He literally posted what I posted, told y'all you were being irrational, and the response was, "he said what E-Balla didn't." No he just did what I told y'all to do and went to basketball-reference for literally 2 seconds before making his opinion on something. If you can't be bothered to even read my post or visit basketball-reference before responding, don't.


Awww, is someone upset another person is stealing his glory?

If you make a case, the burden is on you to support it. If it takes “literally two seconds”, why did you not do it just to back yourself up? Oh, wait, I know, your time is so profoundly valuable that it makes much more sense for each of us to do that work ourselves, right? After all, if you did that, you would lose all this time you devoted to blaming everyone else for not deifying your data.

If you need clarification, ask.


Hahahahaha, I wish I could highlight this further. This has to be self-parody, no one is this deeply out of touch.

Beyond that it's obvious he wasn't even attempting to understand the argument from the jump and you clearly jumped in pissed off over the Curry stuff and didn't even understand what was even being discussed before you jumped in.


Yes, everyone else is the problem, not you. Everyone else is biased. It is all a big conspiracy. We are just maliciously trying to waste your time because we are jealous of your unmatched intellect. You have it all figured out.

Again, get over yourself. If it is you against the world, maybe stop to consider if possibly there could be a legitimate reason.
freethedevil
Head Coach
Posts: 7,262
And1: 3,236
Joined: Dec 09, 2018
         

Re: Peaks project update: #12 

Post#85 » by freethedevil » Wed Aug 7, 2019 6:22 am

E-Balla wrote:
freethedevil wrote:
E-Balla wrote:

Their boxscore stats are great, their BPM is great, their PER is great. Anyone looking at their boxscore numbers could tell you those things already, what's the value in having BPM specifically tell you those things?
.

The value is bpm's weighting isn't entirely arbitrary like, say, per. It uses how well the box #'s correlate with rapm which itself comes from game score. So while far from ideal, it correlates better with winning. And you don't have to take my word for that, this kind of thing's already been tested. BPm does a better job predicting team success than per and is more stable as players leave and go indicating it's less circumstance sensitive.

On the other hand, # of assists usually have little to no correlation with the quality of an offense and ppg is far less stable than even per.

Again but what can be said for having a high BPM? What does it tell you about a player? How does it provide more and not considerably less context than using the boxscore yourself? Because some numbers correlate with winning? According to BPM Lamarcus is a worse offensive player than Corey Maggette. Any single stat claiming to be a ranking system of players is just nonsense.

It estimates impact, quite literally the point of signing basketball players, per minuite. Literally nothing in the boxscore can do that so right off the bat that's more valuable. That's the whole point of rapm's prior. You can take issue with rapm but the results speak for themselves.


High PPG with high efficiency tells me the player is a great scorer does it not? A high BPM tells me... What exactly?[/quote]

According to BPM Lamarcus is a worse offensive player than Corey Maggette.

Large sample size of data says bpm useful. U say "no, maggette", I'm convinced.


Regardless if this is as a ridiuclous conclusions as you claim, then it should be easily avoided with understanding the stat. I'm confident in saying the ridiculousness of this conclusions would come down to two things.

1. "BPM" is per minuite, so a player with a far lesser role can be vastly less valuable than a "less effecient" player with a bigger role. So if corey is a scrub and lamarcus is not, using vorp(accumulative bpm) should tell you otherwise.

2. Secondly, bpm measures impact, but impacting a better team is harder than impacting a worse team. 20% of a skyscraper is bigger than 30% of a hut. To ball park a player's impact, which box score doesn't measure, using bpm, you look at
a. their impact
b. the size of the building they're impacting.

Luck adjusted stats do b for us but they only go so far back. However even common sense can approximate that effect. Here's an example:

Giannis's bpm was a teensy bit below Harden's. But the bucks srs was WAAAY higher than the rocket's. Assuming you're using your head, bpm is useful in telling you that Giannis was nearly impactful as harden. Assuming you're using your head, that he was nearly as impactful for a much better team should tell you Giannis was significantly better.

And guess what?

Instead of using basic box stats or box based stats like the media did to give harden an argument for mvp, if i cross check that with luck adjusted stats that did a far better job predicting results than vegas did and have consistently proven themsleves to be the most stable and predictive of stats, I'll find I'm probably right:
https://www.bball-index.com/18-pipm/

If I used box stats to assess impact, I would have come to the laughable conclusion that harden is better than giannis. Fortuantely i did not. Now.... lets replace harden with a 80's player who scored a lot and was a great playmaker while dominanting the ball on a team that was meh. Suddenly i can't use luck adjusted measures. So should I

a. Use box stats that say player b is better than player a
or
b. Use bpm and srs.

I think the latter choice is the smart one 100/100 times.

High PPG with high efficiency tells me the player is a great scorer does it not? A high BPM tells me... What exactly?

2019 Harden and 2016 Curry were both great scorers. Whose scoring was better?

You might cite the combination of great effiency and great volume to say it was harden but you'd be wrong. Harden's ball dominance diminishes the value of his scoring. Curry being able to score like he does with far less usage of the ball allows for more opportunities to score. Would box stats tell you that?

No.

You know what would?

BPM.

ScoreVal, which uses bpm weighting to judge the value of specific aspects the game tell us the impact of harden's scoring was worse than the impact of curry's which had the highest regular season score by a decent margin.

Box stats hold value as the first step on making qualitative assessments of play, but that's all. Those qualitiative statments can, at best, allow you to look at skillsets and see what's more beneficial to championship calibre teams. However that won't get you anywhere unless you have an idea of impact.
User avatar
E-Balla
RealGM
Posts: 35,818
And1: 25,114
Joined: Dec 19, 2012
Location: The Poster Formerly Known As The Gotham City Pantalones
   

Re: Peaks project update: #12 

Post#86 » by E-Balla » Wed Aug 7, 2019 11:23 am

liamliam1234 wrote:It is really funny, EB, how you are still going on with this. Again, what does it say if your ego is so fragile that you refuse to acknowledge anyone not immediately accepting every little point you make?

E-Balla wrote:If he wanted to gather information basketball-reference is right there and he already knows the criteria being used. I'm not one for the gish gallop.


Yes, totally a gish gallop for me to not want to redo your work from scratch. I know whenever I read any academic study or data analysis, the authors always say, “Hey, if you want to look at some extra variables of our data set, screw off, you are on your own.”

He clearly didn't read my posts, why gather more statistics for him to write off and ignore?


Telling that this is all you can ever say. Maybe we should all follow in your lead. “Oh, you called my brilliance into question? You just must not be reading properly.” And it is especially convenient when you using it as an excuse to willfully misread and misinterpret every post against you. Tell me, does this perpetual sense of victimisation keep you up at night?

E-balla wrote:
Blatantly dismissing metrics because of his “eyes do a better job” lol.

"Metrics". I'll come up with a number called super player rating made by averaging their place on each NBA leaderboard. That's now "metrics" worth discussing no matter how worthless?


Probably would do a better job than the vast majority of eye tests. But as you have incessantly sought to establish, your eye test must naturally be flawless.

E-balla wrote:
I’m not sure if requesting to know the source of the data qualifies as shifting burden of proof.

Find one post prior to this where anyone mentioned needing a source?


Totally a good-faith interpretation of his point. Yes, we all know you “sourced” the information from basketball-reference. Maybe contrast your posts with Frog’s and see if you can figure out the difference. Oh, wait, you are so much of a narcissist you literally cannot.

E-balla wrote:
What’s to stop any/everyone from making up random stats and then forcing the other party to do the math to refute it? Liamliam1234 requested sources and wanted to know specifically which teams/series the stats applied to.

You did a fantastic job of actually presenting further context which does confirm E-balla’s original claim about their peak playoff performances.

Nah he has a patience for bull.

He literally posted what I posted, told y'all you were being irrational, and the response was, "he said what E-Balla didn't." No he just did what I told y'all to do and went to basketball-reference for literally 2 seconds before making his opinion on something. If you can't be bothered to even read my post or visit basketball-reference before responding, don't.


Awww, is someone upset another person is stealing his glory?

If you make a case, the burden is on you to support it. If it takes “literally two seconds”, why did you not do it just to back yourself up? Oh, wait, I know, your time is so profoundly valuable that it makes much more sense for each of us to do that work ourselves, right? After all, if you did that, you would lose all this time you devoted to blaming everyone else for not deifying your data.

If you need clarification, ask.


Hahahahaha, I wish I could highlight this further. This has to be self-parody, no one is this deeply out of touch.

Beyond that it's obvious he wasn't even attempting to understand the argument from the jump and you clearly jumped in pissed off over the Curry stuff and didn't even understand what was even being discussed before you jumped in.


Yes, everyone else is the problem, not you. Everyone else is biased. It is all a big conspiracy. We are just maliciously trying to waste your time because we are jealous of your unmatched intellect. You have it all figured out.

Again, get over yourself. If it is you against the world, maybe stop to consider if possibly there could be a legitimate reason.

It's not me against the world, the whole reason a whole new poster jumped into the fray was to let you guys know you were being completely unreasonable. If anything it's you against the world, and you're still refusing to acknowledge what you wanted from this whole exchange was to discredit my data for unsubstantiated reasons because you didn't like the point of what it said. At no point did you even engage the argument being made.

You keep focusing on some ego thing, it's not about that it's about you asking for statistics that don't exist and calling me biased for not using them. The gap between me and frog's post is obvious, like I said I had no time for the gish gallop, he did. He went and found you guys pace data for the teams. I refused to do something you could find by easily bringing up basketball reference in 2 windows. Maybe you need to learn what gish gallop is...

Gish Gallop is a technique, named after the creationist Duane Gish who employed it, whereby someone argues a cause by hurling as many different half-truths and no-truths into a very short space of time so that their opponent cannot hope to combat each point in real time.


Like for example someone not looking up pace data but using it as an excuse in why there's a gap in production, someone pretending it wasn't well defined what was a good and average defense was that used that to discredit the data, someone that gives you 50 theories for why your data sucks without ever checking the validity of any of their arguments l. He didn't say anything you wouldn't find out after immediately looking at basketball-reference.

Also to the bolded, no but it's totally gish gallop to ask me to completely rerun data for 10 years of the playoffs to get their BPM in those series. You know how I know I gave you enough tools to see those things yourself (if you cared to)? Frogbros found it. Props to him for taking the hour or so it took to write that, unfortunately in a project where I'm already pressed for time I'm not engaging in bad faith arguments. Before you doubt the validity of data given to you I get that it makes sense to ask for data, if the poster says the data doesn't exist, it's no longer reasonable to continue to pester them for that data (another thing frogbros' post pointed out). That's what I meant when I said he said what I said, I'm not talking about the data. I'm talking about him throughly explaining why it was obvious you guys were engaging in this conversation in bad faith to start out.

Also no one asked you to redo data from scratch. Is that what you call visiting basketball reference to look at the pace of the teams? Is that what you call me asking you to read my posts? You claimed I didn't even say what a good and average defense was when I did that immediately before posting the data. Why would anyone see someone do that and think they're posting in good faith or read the post?

Even now after frogbros' whole post about how you guys were arguing in bad faith you still somehow think it's me against the world. That's an unbiased third party coming in to say I wasn't crazy, but your response to it is to completely ignore the existence of every part of the post calling you on your bs.
Mavericksfan
Senior
Posts: 531
And1: 198
Joined: Sep 28, 2011

Re: Peaks project update: #12 

Post#87 » by Mavericksfan » Wed Aug 7, 2019 11:34 am

Liamliam1234 wrote:Totally a good-faith interpretation of his point. Yes, we all know you “sourced” the information from basketball-reference. Maybe contrast your posts with Frog’s and see if you can figure out the difference. Oh, wait, you are so much of a narcissist you literally cannot.


The funny thing you literally asked and were ignored.

But the numbers do not clearly show that because you could not be bothered to provide all the context. I have no idea what series are being selected for Robinson, or how many. I have no idea if pace is a factor. Nor do I have a clear idea what the outside rates are as comparison; if Ewing were just a bit below Robinson against mediocre/bad defences, that would speak to your point a lot more than if Ewing simply had a base level offensive performance regardless of opponent.


His scoring was admittedly garbage (still won 3-1, though). But you know, that is an interesting point. Because by that little average sample you posted, it seemed like Robinson played enough great defences to showcase a clear inferiority against them. So where exactly did those numbers come from? This is what I am talking about with you making broad declarations without bothering to show your work.


First quote is to add the context of you stating you wanted more data than he supplied. I bolded the part where you quite literally requested to know where his numbers are coming from. Obviously that refers to the data set itself not which website.

I removed his name because I won’t be responding to him anymore. He’s shown a willingness to lie and deflect during discussion so I won’t be wasting my time.
User avatar
E-Balla
RealGM
Posts: 35,818
And1: 25,114
Joined: Dec 19, 2012
Location: The Poster Formerly Known As The Gotham City Pantalones
   

Re: Peaks project update: #12 

Post#88 » by E-Balla » Wed Aug 7, 2019 11:55 am

freethedevil wrote:It estimates impact, quite literally the point of signing basketball players, per minuite. Literally nothing in the boxscore can do that so right off the bat that's more valuable. That's the whole point of rapm's prior. You can take issue with rapm but the results speak for themselves.

So you're telling me with no further context you'd feel confident claiming (for example) player B was a +2 player because they had a +2 BPM? That's good basketball analysis to you? That's extra context to you?

Large sample size of data says bpm useful. U say "no, maggette", I'm convinced.

As I said BPM works on wide samples because it's made to ultimately line up in terms of team to team data and large sets of players. Still it doesn't work for player to player comparisons because it can, and commonly is, way off on players. Steve Nash isn't a +1.3 player. Moses Malone isn't a +1.7 player. Using these numbers for player to player comparisons, while not acknowledging they're wildly off on some players is just lazy. Yes Maggette is relevant here, I'm talking about why all all in one stats suck. TS% tells me player A was more efficient scoring than player B that's something factual to cite. All in one stats tell you player A has more impact than player B, but it's not at all factual. Why use it?

Regardless if this is as a ridiuclous conclusions as you claim, then it should be easily avoided with understanding the stat. I'm confident in saying the ridiculousness of this conclusions would come down to two things.

1. "BPM" is per minuite, so a player with a far lesser role can be vastly less valuable than a "less effecient" player with a bigger role. So if corey is a scrub and lamarcus is not, using vorp(accumulative bpm) should tell you otherwise.

Maybe you didn't know this but Maggette was a starter playing full minutes when he got better results than LMA has ever had in OBPM.

2. Secondly, bpm measures impact, but impacting a better team is harder than impacting a worse team. 20% of a skyscraper is bigger than 30% of a hut. To ball park a player's impact, which box score doesn't measure, using bpm, you look at
a. their impact
b. the size of the building they're impacting.

Luck adjusted stats do b for us but they only go so far back. However even common sense can approximate that effect. Here's an example:

Giannis's bpm was a teensy bit below Harden's. But the bucks srs was WAAAY higher than the rocket's. Assuming you're using your head, bpm is useful in telling you that Giannis was nearly impactful as harden. Assuming you're using your head, that he was nearly as impactful for a much better team should tell you Giannis was significantly better.

How does BPM say ANYTHING about his impact? See with modern players we can actually match up their RAPM (aka impact) to their BPM and see in the case of Harden (using him for this example) he's leading the league in BPM yearly while placing 2nd on his own team in the REAL version of the impact number BPM is meant to approximate. Even in your example there's a clear bad conclusion you gather from using that bad data, assuming I'm using my head I'll use RAPM and clearly see Harden isn't approaching Giannis' impact at all. Clearly isn't approaching prime Wade and Kobe's impact at all.

And guess what?

Instead of using basic box stats or box based stats like the media did to give harden an argument for mvp, if i cross check that with luck adjusted stats that did a far better job predicting results than vegas did and have consistently proven themsleves to be the most stable and predictive of stats, I'll find I'm probably right:
https://www.bball-index.com/18-pipm/

You're pretending we HAVE to rely on only the boxscore, as if we can't draw conclusions from other things along with the boxscore, but at the end of the day to draw your whole conclusion you go to a stat that isn't BPM just to confirm what BPM said wasn't accurate, and that's why BPM is useful? Why not jump straight to PIPM?

If I used box stats to assess impact, I would have come to the laughable conclusion that harden is better than giannis. Fortuantely i did not. Now.... lets replace harden with a 80's player who scored a lot and was a great playmaker while dominanting the ball on a team that was meh. Suddenly i can't use luck adjusted measures. So should I

a. Use box stats that say player b is better than player a
or
b. Use bpm and srs.

I think the latter choice is the smart one 100/100 times.

The problem is like I said a million times (and this is another point where another poster jumped in to confirm what I was saying was completely accurate because you refused to acknowledge this) BPM was made with the modern era in mind. In the 80s was the value of offensive rebounding the same as now? Was good 3 point shooting as highly correlated to good offense as it is now? Nope not at all. Why use a stat created to approximate impact in the context of the modern NBA to make conclusions on players NOT in the modern NBA? Why use it for the modern NBA when we have +/- stats doing that just well?

That's before we get into how completely crappy all defensive boxscore estimates are (so half of BPM is just absolute garbage).

2019 Harden and 2016 Curry were both great scorers. Whose scoring was better?

Curry by far. Not even really close. The volume is so insanely far off from each other, we've seen people around Harden's volume but more efficient, never people around Curry's volume and anywhere near his efficiency. That's something you can easily pick up from the boxscore.

You might cite the combination of great effiency and great volume to say it was harden but you'd be wrong. Harden's ball dominance diminishes the value of his scoring. Curry being able to score like he does with far less usage of the ball allows for more opportunities to score. Would box stats tell you that?

No.

Well good thing I don't have to limit myself to the boxscore!

You know what would?

RAPM? NBA tracking data? Watching games to see how they came across those points? BPM isn't telling you that. Actually BPM is telling you Harden is better than Curry outside of 2016. BPM is telling you Harden had his best year last season. BPM is literally just letting you know when he had his gaudiest numbers, not when he had his biggest impact. It just doesn't lineup to the ACTUAL impact data we have on Harden at all.

Actually you've done well using him as you example so often because Harden is a player BPM is always completely wrong on.

ScoreVal, which uses bpm weighting to judge the value of specific aspects the game tell us the impact of harden's scoring was worse than the impact of curry's which had the highest regular season score by a decent margin.

Box stats hold value as the first step on making qualitative assessments of play, but that's all. Those qualitiative statments can, at best, allow you to look at skillsets and see what's more beneficial to championship calibre teams. However that won't get you anywhere unless you have an idea of impact.

Cool, BPM doesn't give me the idea of impact for any individual player. For large sets of players it allows me to approximate their impact in the context of the modern NBA the way the game is played now. Beyond that it's not particularly useful and you basically wrote a whole post about how you CAN'T use it context free to make any point on a player so thanks for proving my point.
liamliam1234
Senior
Posts: 679
And1: 663
Joined: Jul 24, 2019

Re: Peaks project update: #12 

Post#89 » by liamliam1234 » Wed Aug 7, 2019 11:58 am

E-Balla wrote:It's not me against the world, the whole reason a whole new poster jumped into the fray was to let you guys know you were being completely unreasonable.


He jumped into the fray to do what you were refusing to do. Or did you miss the part where discussion immediately improved after he did that? No, I guess you probably would not.

E-Balla wrote:If anything it's you against the world, and you're still refusing to acknowledge what you wanted from this whole exchange was to discredit my data for unsubstantiated reasons because you didn't like the point of what it said. At no point did you even engage the argument being made.


A.) Lol, and the deluded persecution complex swings right back in.

B.) Because you do such a good job helping us engage, right.

E-Balla wrote:You keep focusing on some ego thing, it's not about that it's about you asking for statistics that don't exist and calling me biased for not using them. The gap between me and frog's post is obvious, like I said I had no time for the gish gallop, he did. He went and found you guys pace data for the teams. I refused to do something you could find by easily bringing up basketball reference in 2 windows. Maybe you need to learn what gish gallop is...

Gish Gallop is a technique, named after the creationist Duane Gish who employed it, whereby someone argues a cause by hurling as many different half-truths and no-truths into a very short space of time so that their opponent cannot hope to combat each point in real time.


Again speaks to your persecution complex if you sincerely believe that is what was going on. You still have yet to entertain the possibility that this was not some conspiratorial act of malicious misinterpretation.

E-Balla wrote:Like for example someone not looking up pace data but using it as an excuse in why there's a gap in production, someone pretending it wasn't well defined what was a good and average defense was that used that to discredit the data, someone that gives you 50 theories for why your data sucks without ever checking the validity of any of their arguments.


More deluded sense of persecution. Again, especially comical given the wailing over people not reading. The only one here who clearly has showed they do not bother to even try to understand the other side is you.

E-Balla wrote:He didn't say anything you wouldn't find out after immediately looking at basketball-reference.


Why ever cite to anything if it is derived from basketball-reference, right. You know you may as well have just told us that Robinson was worse offensively against non-bad playoff defences, without giving any numbers, because “it was all on basketball-reference”.

E-Balla wrote:Also to the bolded, no but it's totally gish gallop to ask me to completely rerun data for 10 years of the playoffs to get their BPM in those series.


Feel free to show me where I asked you to manually calculate their BPM.

E-Balla wrote:You know how I know I gave you enough tools to see those things yourself (if you cared to)? Frogbros found it.


Yes, thank you for “giving” us the tools of... basketball reference. So generous. Such a valid interpretation of the issue.

E-Balla wrote:Props to him for taking the hour or so it took to write that, unfortunately in a project where I'm already pressed for time I'm not engaging in bad faith arguments.


How many hours have you wasted on these contentless replies.

E-Balla wrote:Before you doubt the validity of data given to you I get that it makes sense to ask for data, if the poster says the data doesn't exist, it's no longer reasonable to continue to pester them for that data (another thing frogbros' post pointed out).


So we were unreasonable to ask for both nonexistent data and existing data we could access by doing the work to replicate everything you did. In other words, we were unreasonable for asking for data.

E-Balla wrote:That's what I meant when I said he said what I said, I'm not talking about the data. I'm talking about him throughly explaining why it was obvious you guys were engaging in this conversation in bad faith to start out.


Frogbros can join in if he wants, and I encourage him to, but for my reading there was nothing to remotely suggest he thought it was an issue of “bad faith”. More self-absorbed delusion.

E-Balla wrote:Also no one asked you to redo data from scratch.


No, it was just implicit in your refusal to do anything or give anything or share anything.

E-Balla wrote:Is that what you call visiting basketball reference to look at the pace of the teams?


Gee, I missed the part where Frogbro’s post just showed your numbers per 100.

E-Balla wrote:Is that what you call me asking you to read my posts?


Gee, I missed the part where Frogbros literally just copied your post. Oh, wait, I forgot you still think that is basically all he did.

E-Balla wrote:You claimed I didn't even say what a good and average defense was when I did that immediately before posting the data.

Why would anyone see someone do that and think they're posting in good faith or read the post?


Oh, look, I was right, you never figured this one out.

Alright, I guess I had my laugh. I said you never specified a “great” defence, and I did that after you specifically said the -2.8 Blazers did not qualify as a “great” defence. Again, for someone supposedly so concerned with reading, you sure do consistently do the worst possible job of it.

E-Balla wrote:Even now after frogbros' whole post about how you guys were arguing in bad faith you still somehow think it's me against the world.


Amazing how one person taking pity on you and deciding to do the work you stubbornly refused to do has gone completely to your head. But again, not surprising from an unrepentant egoist.

E-Balla wrote:That's an unbiased third party coming in to say I wasn't crazy, but your response to it is to completely ignore the existence of every part of the post calling you on your bs.


1. Because people are only unbiased when they agree with you, right.

2. Maybe you missed the part where we all acknowledged he gave us what he had been requesting. Yet another surprise from someone so full of himself that he reads neutral clarification as a blatant condemnation of opponents.
User avatar
E-Balla
RealGM
Posts: 35,818
And1: 25,114
Joined: Dec 19, 2012
Location: The Poster Formerly Known As The Gotham City Pantalones
   

Re: Peaks project update: #12 

Post#90 » by E-Balla » Wed Aug 7, 2019 12:02 pm

Mavericksfan wrote:
Liamliam1234 wrote:Totally a good-faith interpretation of his point. Yes, we all know you “sourced” the information from basketball-reference. Maybe contrast your posts with Frog’s and see if you can figure out the difference. Oh, wait, you are so much of a narcissist you literally cannot.


The funny thing you literally asked and were ignored.

But the numbers do not clearly show that because you could not be bothered to provide all the context. I have no idea what series are being selected for Robinson, or how many.

I'll stop it right here, how did frogbros go easily look if I didn't give you everything you needed to figure out what series were being selected? I literally told you guys the level of defense being considered at each level, from there figuring out what series is as easy as going to basketball reference. If he had no idea of how many series were being used for Robinson, even after being given the criteria for what series were being used, that shows he clearly doesn't care to know.

Also none of that is asking for a source so...
liamliam1234
Senior
Posts: 679
And1: 663
Joined: Jul 24, 2019

Re: Peaks project update: #12 

Post#91 » by liamliam1234 » Wed Aug 7, 2019 12:09 pm

No one is saying it was impossible to do the work. You are still missing the point, and you still have absolutely no grasp of the expectations on someone ostensibly trying to “defend” a specific idea. Maybe because you see everything as a personal attack. It is not the job of every individual audience member to recreate the work; if that element of presentation is too large a burden for you, you may as well just acknowledge you want to post things once and be done with it.
User avatar
E-Balla
RealGM
Posts: 35,818
And1: 25,114
Joined: Dec 19, 2012
Location: The Poster Formerly Known As The Gotham City Pantalones
   

Re: Peaks project update: #12 

Post#92 » by E-Balla » Wed Aug 7, 2019 12:23 pm

liamliam1234 wrote:Again speaks to your persecution complex if you sincerely believe that is what was going on. You still have yet to entertain the possibility that this was not some conspiratorial act of malicious misinterpretation.

Yes or no, when you made your first post, I said that data doesn't exist, and you asked for it again saying my data was completely unreliable without it, how was I supposed to interpret that? When you never engaged my argument but looked for 50 million ways to discredit the data without taking 2 seconds to look at if your refutations had any base in reality before typing them how was I supposed to interpret that? Maybe you do it so much you didn't realize those are bad faith arguing tactics, but they are. If I read a post and want to discredit a piece of data in it with a point, I've literally never done so without first visiting basketball reference to see if I'm about to put my foot in my mouth. Why? I appreciate other people's time. I appreciate the work that goes into good posts. I don't appreciate people adding nothing to the conversation and lazily discrediting data off of hunches.

Feel free to show me where I asked you to manually calculate their BPM.

When I said it doesn't exist and you kept pressing me on posting it? Was that not you telling me to go manually calculate it?

How many hours have you wasted on these contentless replies.

Way less than it takes for me to make a single good post because I research everything I say before saying it. I have a post on Moses Malone for the next thread I've been working on for 2 days now in Evernote. Maybe for you who just speed stuff posts like these are the most time you take to post, but I know that's not true because outside of this particular conversation you've been good at using data to substantiate your points.

So we were unreasonable to ask for both nonexistent data and existing data we could access by doing the work to replicate everything you did. In other words, we were unreasonable for asking for data.

The unreasonable part was that you discredited the data immediately and only the inclusion of that data would make you credit it. You didn't make a post refuting my point. You made a point questioning my data, with no real reason, and not at all addressing my point. You (or was it Mavs fan) even said after frogbros spent the time to gather that data that only then was your opinion changed. So quite literally you (or him, whoever it was) refused to even take into account the validity of my argument before even going to make sure you had any reason to do so. That's bad faith arguing 1.0.

I found this on google real quick:

https://www.reddit.com/r/AskALiberal/comments/b3zk8l/rule_5_refresher_how_to_argue_in_good_faith/

5. Do not dismiss evidence out of hand. If provided with evidence of a claim you that you reject, it is not considered good faith to refuse to give that source any credence without sufficient evidence to do so.


Frogbros can join in if he wants, and I encourage him to, but for my reading there was nothing to remotely suggest he thought it was an issue of “bad faith”. More self-absorbed delusion.


From his post if you bothered to read anything before he posted the data:

E-Balla made a very specific claim: Ewing performed better offensively in the playoffs against higher quality defenses throughout their entire primes. E-Balla presented enough evidence to at least support this claim on the ground level. LiamLiam's retort to this was to disagree with the idea in general because it somehow didn't pass his sniff test, and proceeded to throw out BPM and Winshare #'s for 2 single, different seasons for each player (90 Ewing and 95 Robinson).

How is citing regular season BPM and Winshares any type of refutation that Ewing played better against tougher defenses in the playoffs over the course of their entire respective primes? How is citing playoff BPM and Winshares for 2 single, different playoff runs a refutation that Ewing played better against tougher defenses in the playoffs over the course of their entire respective primes? It's not.


He has said multiple times that the specific data you may be insisting on that would actually provide further clarification does not exist. Since he has at least already presented SOME evidence (as opposed to none), at what point does the burden of proof shift to the party disagreeing to refute the evidence that WAS presented? If you want to simply disagree with the evidence he provided, that's fair. But it seems a bit unrealistic to clamor about for additional, custom requested pieces of data to satisfy whatever qualms you have about the claim being made.


He might've not used the term "bad faith" but he certainly described it. Maybe you had no idea what bad faith arguments are and no idea you were engaging in a bad faith argument but think one thing to yourself, when you originally responded to my post was the purpose to add to the discussion, or was it to discredit the statistics I posted. And be honest with yourself here, because you didn't add anything to the conversation, but you certainly did attempt to discredit my numbers so if you were trying to add to the discussion you fumbled that.

Alright, I guess I had my laugh. I said you never specified a “great” defence, and I did that after you specifically said the -2.8 Blazers did not qualify as a “great” defence. Again, for someone supposedly so concerned with reading, you sure do consistently do the worst possible job of it.

No but I specified an average (-2 to +2) and good (-2 to -4 defense) which is a range including the Blazers so anyone reading that should've seen they're a good defense, not what falls into my criteria of great (I think I used the term elite but still). That's a well defined criteria there.
User avatar
E-Balla
RealGM
Posts: 35,818
And1: 25,114
Joined: Dec 19, 2012
Location: The Poster Formerly Known As The Gotham City Pantalones
   

Re: Peaks project update: #12 

Post#93 » by E-Balla » Wed Aug 7, 2019 12:26 pm

liamliam1234 wrote:No one is saying it was impossible to do the work. You are still missing the point, and you still have absolutely no grasp of the expectations on someone ostensibly trying to “defend” a specific idea. Maybe because you see everything as a personal attack. It is not the job of every individual audience member to recreate the work; if that element of presentation is too large a burden for you, you may as well just acknowledge you want to post things once and be done with it.

No one asked you to recreate my work. What's being asked is to not discredit my work with no reason to. If you want to discredit someone's hard work, put your own work in. Like I said I respect people's times I'm never looking at a post where someone ran the numbers and discrediting it off of hunches that may or may not lead to anything and asking them to do the legwork to see if my hunches are even worth the hours they're going to spend chasing them down and typing them into a format that's easily readable and digestible, ESPECIALLY when I didn't take the time to at least address their point, their argument, or understand what their post even said.
liamliam1234
Senior
Posts: 679
And1: 663
Joined: Jul 24, 2019

Re: Peaks project update: #12 

Post#94 » by liamliam1234 » Wed Aug 7, 2019 12:53 pm

Rather than respond point by point again, I think the main issue is that you felt there was an intent to discredit your work. There was not. Alright, at the start I could have been more considered in my criticism, sure – although in terms of ill-considered tone, that is hardly a one-way street – but my issue was that the data was so specific that no one could create a counter-point without redoing your work themselves (and although you keep saying that would not have been the case, considering you effectively wanted us to work backward and figure out which defences applied to the samples year by year, and then calculate whatever we wanted to after doing that, I would say at that point it may as well be a replication of the work you did). I was not dismissing it because all I had was what you gave. But your data is not the whole story, and I think you understand that. And you making a case and then refusing to follow up is, as we have seen, an excellent recipe to kill discussion, because then the issue becomes the lack of alternate data rather than the data itself.

Also, bad faith debating is fundamentally and strictly about intent, so no, that is not what was at play. Again, the intent was never to dismiss; the intent was to analyse the framing, but you had no interest to helping people do that. And alright, I can understand not wanting to spend the extra time, but there was a much better way to go about that than the route you took.

Also also, I offered the possibility of attempted some average BPM (or whatever) per series given how the bulk of their runs only went two rounds, and how the several which ended in the first round would obviously have those metrics already calculated. That is pretty beside the point now, though. I just wanted extra data to analyse, but that did not happen until someone else took over on your behalf.

Also also also, do you at least see how you confusingly conflated great/elite with good?
User avatar
E-Balla
RealGM
Posts: 35,818
And1: 25,114
Joined: Dec 19, 2012
Location: The Poster Formerly Known As The Gotham City Pantalones
   

Re: Peaks project update: #12 

Post#95 » by E-Balla » Wed Aug 7, 2019 3:32 pm

liamliam1234 wrote:Rather than respond point by point again, I think the main issue is that you felt there was an intent to discredit your work. There was not. Alright, at the start I could have been more considered in my criticism, sure – although in terms of ill-considered tone, that is hardly a one-way street – but my issue was that the data was so specific that no one could create a counter-point without redoing your work themselves (and although you keep saying that would not have been the case, considering you effectively wanted us to work backward and figure out which defences applied to the samples year by year, and then calculate whatever we wanted to after doing that, I would say at that point it may as well be a replication of the work you did). I was not dismissing it because all I had was what you gave. But your data is not the whole story, and I think you understand that. And you making a case and then refusing to follow up is, as we have seen, an excellent recipe to kill discussion, because then the issue becomes the lack of alternate data rather than the data itself.

No an excellent way to kill discussion is to immediately dismiss a statistic in the first place for no good reason at all.

If you remember your original post, while oddly confrontational and dismissive, got a response where I just asked what advanced stats you wanted to see from 88 and 90 that would paint a clearer picture. Your response to that question was win shares and BPM. Those don't exist and I said that. At that point it should've been dropped as a point but again you clearly didn't have good intentions so your next post was you attempting to derail the conversation, ignore what I posted, and deflect whenever I attempted to get you to address my post.

You're at this moment either arguing in bad faith now by saying you read my post before the original response (sidebar: When I say read my post I don't mean literally see the words with your eyes, I mean understand them and the point being made), or you were arguing in bad faith from the jump. Would you at least admit now, 2 threads later, that at no point did you actually engage ANYTHING my post said and you instead questioned the validity of the data?

I was going to address the rest of your post but honestly if you can't admit you didn't have good intentions in that first post there's no point. If that's a post made in good faith what the hell does one made in bad faith look like?
liamliam1234
Senior
Posts: 679
And1: 663
Joined: Jul 24, 2019

Re: Peaks project update: #12 

Post#96 » by liamliam1234 » Wed Aug 7, 2019 4:22 pm

E-Balla wrote:
liamliam1234 wrote:Rather than respond point by point again, I think the main issue is that you felt there was an intent to discredit your work. There was not. Alright, at the start I could have been more considered in my criticism, sure – although in terms of ill-considered tone, that is hardly a one-way street – but my issue was that the data was so specific that no one could create a counter-point without redoing your work themselves (and although you keep saying that would not have been the case, considering you effectively wanted us to work backward and figure out which defences applied to the samples year by year, and then calculate whatever we wanted to after doing that, I would say at that point it may as well be a replication of the work you did). I was not dismissing it because all I had was what you gave. But your data is not the whole story, and I think you understand that. And you making a case and then refusing to follow up is, as we have seen, an excellent recipe to kill discussion, because then the issue becomes the lack of alternate data rather than the data itself.

No an excellent way to kill discussion is to immediately dismiss a statistic in the first place for no good reason at all.

If you remember your original post, while oddly confrontational and dismissive, got a response where I just asked what advanced stats you wanted to see from 88 and 90 that would paint a clearer picture. Your response to that question was win shares and BPM. Those don't exist and I said that. At that point it should've been dropped as a point but again you clearly didn't have good intentions so your next post was you attempting to derail the conversation, ignore what I posted, and deflect whenever I attempted to get you to address my post.

You're at this moment either arguing in bad faith now by saying you read my post before the original response (sidebar: When I say read my post I don't mean literally see the words with your eyes, I mean understand them and the point being made), or you were arguing in bad faith from the jump. Would you at least admit now, 2 threads later, that at no point did you actually engage ANYTHING my post said and you instead questioned the validity of the data?

I was going to address the rest of your post but honestly if you can't admit you didn't have good intentions in that first post there's no point. If that's a post made in good faith what the hell does one made in bad faith look like?


Christ, it is like every chance you have to achieve a sliver of self-awareness you just throw it out the window. Now you are at the point of arrogance to tell me my own intentions (while also, surprise surprise, still what at this point has to be deliberately misreading the entire conversation). Alright, how is this for bad intent: if you only post to foster your own colossal ego, without ever considering the possibility that you may have misinterpreted or been excessively stubborn or even not given the definitive take on the subject, you are starting the discussion in bad faith and should not be taken seriously. I have more patience for this type of smugly unctuous dreck than most, but I am really starting to envy those who bailed as soon as they realised you will never stop acting as if everyone else is beneath you.
User avatar
E-Balla
RealGM
Posts: 35,818
And1: 25,114
Joined: Dec 19, 2012
Location: The Poster Formerly Known As The Gotham City Pantalones
   

Re: Peaks project update: #12 

Post#97 » by E-Balla » Wed Aug 7, 2019 4:52 pm

liamliam1234 wrote:
E-Balla wrote:
liamliam1234 wrote:Rather than respond point by point again, I think the main issue is that you felt there was an intent to discredit your work. There was not. Alright, at the start I could have been more considered in my criticism, sure – although in terms of ill-considered tone, that is hardly a one-way street – but my issue was that the data was so specific that no one could create a counter-point without redoing your work themselves (and although you keep saying that would not have been the case, considering you effectively wanted us to work backward and figure out which defences applied to the samples year by year, and then calculate whatever we wanted to after doing that, I would say at that point it may as well be a replication of the work you did). I was not dismissing it because all I had was what you gave. But your data is not the whole story, and I think you understand that. And you making a case and then refusing to follow up is, as we have seen, an excellent recipe to kill discussion, because then the issue becomes the lack of alternate data rather than the data itself.

No an excellent way to kill discussion is to immediately dismiss a statistic in the first place for no good reason at all.

If you remember your original post, while oddly confrontational and dismissive, got a response where I just asked what advanced stats you wanted to see from 88 and 90 that would paint a clearer picture. Your response to that question was win shares and BPM. Those don't exist and I said that. At that point it should've been dropped as a point but again you clearly didn't have good intentions so your next post was you attempting to derail the conversation, ignore what I posted, and deflect whenever I attempted to get you to address my post.

You're at this moment either arguing in bad faith now by saying you read my post before the original response (sidebar: When I say read my post I don't mean literally see the words with your eyes, I mean understand them and the point being made), or you were arguing in bad faith from the jump. Would you at least admit now, 2 threads later, that at no point did you actually engage ANYTHING my post said and you instead questioned the validity of the data?

I was going to address the rest of your post but honestly if you can't admit you didn't have good intentions in that first post there's no point. If that's a post made in good faith what the hell does one made in bad faith look like?


Christ, it is like every chance you have to achieve a sliver of self-awareness you just throw it out the window. Now you are at the point of arrogance to tell me my own intentions (while also, surprise surprise, still what at this point has to be deliberately misreading the entire conversation). Alright, how is this for bad intent: if you only post to foster your own colossal ego, without ever considering the possibility that you may have misinterpreted or been excessively stubborn or even not given the definitive take on the subject, you are starting the discussion in bad faith and should not be taken seriously. I have more patience for this type of smugly unctuous dreck than most, but I am really starting to envy those who bailed as soon as they realised you will never stop acting as if everyone else is beneath you.

You're doing all these personal attacks. I don't know you, you don't know me. What you're calling smugness is a reaction to something you did is it not? Yes or no when you originally posted did you have the intention of attempting to discredit my numbers or of addressing the conversation and point being made? If it was the latter, why did you do the former? When I said those numbers didn't exist, why didn't you drop there if your goal wasn't to discredit the numbers but instead to engage in a discussion on the information put in the post?
liamliam1234
Senior
Posts: 679
And1: 663
Joined: Jul 24, 2019

Re: Peaks project update: #12 

Post#98 » by liamliam1234 » Wed Aug 7, 2019 5:11 pm

In order: total non-sequitur, no and yes, I didn’t, because as usual you overlooked or misapplied the core of the point.

We are like twenty-five posts in and you have yet to even attempt to understand the other side. Stop projecting and actually try to engage for once. It evidently does not matter if other people try to spell it out for you, so you need to start figuring it out for yourself.

Return to Player Comparisons