Announcement

Collapse
No announcement yet.

Everything Gary Trent Jr.

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • chris wrote: View Post

    what happened to mikal bridges?
    oh shoot I keep mixing the two of them up.

    Comment


    • The Great One wrote: View Post

      Do you really think Trent is in the same level as those guys? bahahaha.
      lol yes unfortunately it appears to be the case... all while ignoring our best defensive line up and the 100th percentile defensive line up in the NBA since January was one that didn't include him in it.

      Comment


      • The Great One wrote: View Post

        Do you really think Trent is in the same level as those guys? bahahaha.
        Wasn't the one who said that.
        It's in that NBA article.

        Comment


        • Kagemusha wrote: View Post

          Wasn't the one who said that.
          It's in that NBA article.
          I think you are misrepresenting the purpose of the article.. again its like how espn has power rankings nba does the same thing to drive their own content as well... this isn't saying ....

          Comment


          • The Claw Reborn wrote: View Post

            So wtf would be the most Reliable and Accurate/ “not noise” statistical measurement for defense then aside from the: Dan I know it all data?
            Impact stats are the best we have. Within a team concept, on/offs are helpful, both with the 4 factors and from an overall DRTG perspective. Matchup stats can be interesting, to evaluate who each player is assigned to, though the success rates are again only helpful in extreme cases.

            And all those stats have drawbacks and limitations, defence is very hard to measure.

            But DFG% is fundamentally noise. There's just not much value there at all even in the context of generally difficult to measure defensive performance.
            twitter.com/dhackett1565

            Comment


            • TrueTorontoFan wrote: View Post

              Now specifically to address this.. I am curious though... I didn't talk about rim protection I gave a link to all of the numbers... whether it was at the rim or 10 feet from the basket or 6 feet from the basket... 3 pointers and 2 pointers.. it gave all it all....
              I know you gave it all. I am pointing out that the only version of DFG% that has been shown to have any predictive value is right at the rim, and even then only in the very very good and very very bad ends of the scale, with most players in the middle bouncing around at random from one year to the next. Where as with, say, midrange DFG%, or 3 point DFG%, even the guys who are extremely good one year are a complete toss up the next - the non-rim versions just mathematically have no value. They describe what happened, but without any predictive value at all it is very hard to ascribe those numbers to actual talent or defensive impact.

              I think it would be fair to look at 3 pointer DFG and 2 pointer DFG and not completely throw it out... I am looking at OG's stats now for the past few years.... what I am looking for and what I was looking for wasn't a specific number but if it was consistently positive or negative overall.... ideally ignoring a rookie season although you can toss that in if you want. Then I looked at Fred's. In both cases I tracked back as far as I could. Again the biggest variance seemed to occur between the first and second year of a players career so I ignored the first year and just looked for the overall trend of positive or negative differentials rather than the exact FG%. AKA did they impact the stat... keep in mind this is over a season(s) worth of duration. Also keep in mind the players I am comparing him to have routinely played on the same squad for their entire career and largely with the same coach as well. I can understand why that would be of concern though because sure some coaches may have certain 'principles'
              You can track it however you like. Any pattern you find is noise. I wish it weren't the case, more useful data is great, but DFG% is really just noise and every player you find that looks like they have a meaningful pattern is very likely just luck of the draw.


              Also keep in mind I went month by month before. Not surprisingly, one of GTJ's best seasons was the bubble season, when he psuedo gained a reputation for being a 3 and D wing because of how he defended in the bubble. That was actually significantly better. You could say that is the one season that us concerning from a stat stand point because ... the bubble took up a chunk of the end of the season so removing that it starts looking no bueno. NBA as a website for the stats isn't good at seperating the raptor vs portland stats within a season.... so if you want even to remove those data points that its fine, however this year, compared to OG, Siakam, Fred..Precious and Barnes you can remove Barnes for the rookie year thing), Gary was the only one who had a positive differential % almost entirely across the board and if you want to remove precious because you only have two years of data the difference looks just as stark.
              That's great! In this instance the largely random number actually fit what we observe in Gary's defence!

              Is it a noisy stat? perhaps I dont know.... I am not sure what you are basing that on specifically .... our players for the most part switch 1-5 including fred..... only one who ... PER CRAFTEDNBA doesn't switch as much based on his versatility score... and the eye test is GTJ.

              Again I am looking for the overall trend not the specific numbers which I think is the right way to utilize this specific statistic... I am not sure if you are suggesting it is useless or not but that would be how I would personally interpret.
              A few years ago I took various DFG% values over several seasons for the entire league and tried to plot predictiveness of the values. Good defensive players should be good every year, or better than the bad defensive players from the previous season, on average anyway. None of that holds up with DFG%. Meaning it is mostly describing how lucky or unlucky the defensive player was, not how good they were at defence.

              It's simply not a useful stat. I'm sorry, because it does support my view of Gary's defence. But it's just not.
              twitter.com/dhackett1565

              Comment


              • As far as I can tell GTJ graded out last season as a positive on defense in almost every defensive advanced stat, except 538 RAPTOR (which is a dubious stat, tbh).

                Comment


                • DanH wrote: View Post

                  Impact stats are the best we have. Within a team concept, on/offs are helpful, both with the 4 factors and from an overall DRTG perspective. Matchup stats can be interesting, to evaluate who each player is assigned to, though the success rates are again only helpful in extreme cases.

                  And all those stats have drawbacks and limitations, defence is very hard to measure.

                  But DFG% is fundamentally noise. There's just not much value there at all even in the context of generally difficult to measure defensive performance.
                  There is not much validity there Dan but carry on what you believe is noise.

                  Comment


                  • golden wrote: View Post
                    As far as I can tell GTJ graded out last season as a positive on defense in almost every defensive advanced stat, except 538 RAPTOR (which is a dubious stat, tbh).
                    Year three for Gary Trent Jr. as a Raptor, the Raw Numbers of this 23 years old has improved year after year….if he ain’t get traded, he is a vital piece to play along side Scottie moving forward.

                    Comment


                    • I also am little curious Dan….you seem to overextend and get out more often with others, lately with TTF but never seen you dispute T-Rex with his flawed arguments about Trent Jr.
                      …or you are just personally encouraged when he likes your posts and tell you to preach your so called noise?

                      Maybe I’m wrong and you just have the same thought process, nothing wrong with that I guess

                      Comment


                      • DanH wrote: View Post

                        I know you gave it all. I am pointing out that the only version of DFG% that has been shown to have any predictive value is right at the rim, and even then only in the very very good and very very bad ends of the scale, with most players in the middle bouncing around at random from one year to the next. Where as with, say, midrange DFG%, or 3 point DFG%, even the guys who are extremely good one year are a complete toss up the next - the non-rim versions just mathematically have no value. They describe what happened, but without any predictive value at all it is very hard to ascribe those numbers to actual talent or defensive impact.



                        You can track it however you like. Any pattern you find is noise. I wish it weren't the case, more useful data is great, but DFG% is really just noise and every player you find that looks like they have a meaningful pattern is very likely just luck of the draw.




                        That's great! In this instance the largely random number actually fit what we observe in Gary's defence!



                        A few years ago I took various DFG% values over several seasons for the entire league and tried to plot predictiveness of the values. Good defensive players should be good every year, or better than the bad defensive players from the previous season, on average anyway. None of that holds up with DFG%. Meaning it is mostly describing how lucky or unlucky the defensive player was, not how good they were at defence.

                        It's simply not a useful stat. I'm sorry, because it does support my view of Gary's defence. But it's just not.
                        look you may be right about it being noise.. you said that you plotted it ... but specifically as a predictive value... which is different in terms of how you look at things statistically.


                        I think when you try to use stats predictively you need to use it in a different manner.. I also don't know what your calculation was or how you tried to limit the so called noise. Did you use regression analysis what were your confounding variables.

                        Again you may be right. I think part of the issue is you used it as a predictive measure.. which I did not. The trend with Trent was spot on... the trend with OG, and the others minus precious and Barnes because of the reasons above.

                        I am not here to say that my "methodology" is right or wrong compared to yours. I used impact stats before and you didn't say a peep. In fact when others refuted the impact stats not specifying one user over another but you didn't appear to chime in even though the impact stats were what they were. I used them.

                        I also don't use impact stats predictively. I believe some do but I have never seen anyone here attempt to predictively guess stats unless its one of those things at the beginning of the season where we all predict the stats of a player.... in which case I still haven't seen anyone try to translate that.



                        My point here is again you may be right you may not be. That said the backwards trend of positive and negative FG differentials OVERALL seems to paint a consistent picture of strong defensive teams when we have had it for the raptors. I did not go team by team and search for every player. I just went to Gobert as an example to make sure I was interpreting the differential correctly which I was.


                        I am aware that defensively we are still learning what stats to use and not use... and really how to even create measures since its still somewhat unknown.... but I do feel as though you are not really understanding what I was looking for and how I was looking at it. Instead you ... and I don't say this with hostility... projected your own exercise and experience with the stat and assumed I did the same. Perhaps... and this is just a PERHAPS thing.. you using it predictively was the ultimate mistake? Is that remotely a possibility that is even .... plausible?



                        I went on to look at Demar today as a good example... he was never seen as a good defender... if it was all noise wouldn't there be some years where OVERALL the differential percentage would largely oscillate? 17-18 and his rookie season 13-14 were the only years that he had an overall negative impact, especially one in multiple categories... some years he had a negative differential for 3 pointers... so I would have to go back and watch the film to see if his closeouts were better then... but overall the other categories seemed consistent.


                        Then I went to look at Draymond because Gobert ..we know what the numbers should be so Draymond defends in both places. Again not focused just on the overall but looked down at each category to see out of 5 how many were negative or positive... and yes coaching and defensive schemes can play a roll in this which is why I chose not to look at players on teams like Utah.... but yeah .... Draymond the trend fits.. EVERY year the differential was negative in all categories I think this year he had a barely positive 3-point value at 0.4 but again I am looking overall not at specific numbers.


                        I do not think it is useless but rather a stat that isn't necessarily used for predictive analysis. Predictive analysis is distinctly unique and different.... an adjacent field so to speak. All stats are past-oriented but only some have qualities that make them useful for future predictions as descriptors.

                        The way I was and am looking at this is as an overall trend that seems to fit... basically every single player I have looked at. It is possible that this is a random anomaly....I am not claiming any causal relation.... I just simply and noticing SO FAR there is a correlation... regardless of the exact relationship and it seems to be a fairly consistent one.. Like I said I could be lucky to so happen to find ones where it all fit. More than willing to admit that. Until I put all the values into a proper spreadsheet (don't have the time) or data frame.


                        What I would be curious to see .. unless you have or are willing to do it.. is how both the overall numbers as well as even the number of positive/negative categories relate to your average impact stat (D-LEBRON/DDARKO). If that has a coefficient of determination that is close to or near zero yeah then I would agree its a purely useless stat. Currently, I don't have the time for that and haven't seen anyone attempt that. If they did they should write an article. It would be interesting.





                        TLDR - without knowing all of the iterations of you looking into the stat... I am assuming you look at it only predictively because of what you stated. I could be wrong in assuming that. For that reason I hesitate to say it is outright useless. However I can understand if you looked at it only in a predictive manner that yes it would be useless entirely.

                        Comment


                        • golden wrote: View Post
                          As far as I can tell GTJ graded out last season as a positive on defense in almost every defensive advanced stat, except 538 RAPTOR (which is a dubious stat, tbh).
                          Just looking at his DBPM yes he was positive for the first time in his career this year, and added defensive winshares for the first time in his career. Gary is on a team with a lot of defensive talent. That said his D-LEBRON rating from where I am looking was negative as was his D-DARKO... as was his CRAFTEDNBA DPM which standardizes across the different metrics on offence with their CRAFTED OPM and on defence with their CRAFTED DPM. He has a negative crafted as well... but the overall score puts him in the 27th percentile.


                          I said this before to someone else.. it could be that crafted pulls their D-LEBRON from a weird place... but at the same time their DBPM and BPM scores in general seem to be the same as basketball reference.

                          Comment


                          • TrueTorontoFan wrote: View Post

                            Just looking at his DBPM yes he was positive for the first time in his career this year, and added defensive winshares for the first time in his career. Gary is on a team with a lot of defensive talent. That said his D-LEBRON rating from where I am looking was negative as was his D-DARKO... as was his CRAFTEDNBA DPM which standardizes across the different metrics on offence with their CRAFTED OPM and on defence with their CRAFTED DPM. He has a negative crafted as well... but the overall score puts him in the 27th percentile.


                            I said this before to someone else.. it could be that crafted pulls their D-LEBRON from a weird place... but at the same time their DBPM and BPM scores in general seem to be the same as basketball reference.
                            We play a different kind of defense.
                            If those metrics are anchored on the way that the rest of NBA plays defense, how can we effectively measure our players then?

                            Comment


                            • TrueTorontoFan wrote: View Post

                              Just looking at his DBPM yes he was positive for the first time in his career this year, and added defensive winshares for the first time in his career. Gary is on a team with a lot of defensive talent. That said his D-LEBRON rating from where I am looking was negative as was his D-DARKO... as was his CRAFTEDNBA DPM which standardizes across the different metrics on offence with their CRAFTED OPM and on defence with their CRAFTED DPM. He has a negative crafted as well... but the overall score puts him in the 27th percentile.


                              I said this before to someone else.. it could be that crafted pulls their D-LEBRON from a weird place... but at the same time their DBPM and BPM scores in general seem to be the same as basketball reference.
                              GTJ is also a positive using D-EPM and DRPM (huge positive, so a bit suspect).

                              Comment


                              • golden wrote: View Post

                                GTJ is also a positive using D-EPM and DRPM (huge positive, so a bit suspect).
                                where do you find EPM?

                                Comment

                                Working...
                                X