MavsBoard

Let's go!!!

https://twitter.com/TheSteinLine/status/...6712726537

(01-01-2022, 12:59 PM)mavsluvr Wrote: [ -> ]Well, if you want to go in that direction -- 

With Luka, the Mavs are winning 57% of their games (more than the 53% that opponent winning percentage would predict).

Without Luka, the Mavs are winning 36% of their games (less than the 43% that the opponent winning percentage would predict). 


My point isn't that people can't cobble together some stats to try to make it look like Luka isn't contributing much, or might even be dragging this band of merry men down. It is that even the stats being offered don't actually show that, the whole premise doesn't pass the smell test, and the most important thing is that the Mavs are a winning team with Luka and a losing team without him. 

Haters gonna hate. (Not accusing you in particular, BTSO).

My point was that stats are typically hard(ish) and wins-x-losses is also one of those stats and not some touchstone that one can turn to when one is confronted with a different stat (whether or not that stat is telling one something). 

(For what it's worth the Mavs "expected" win %ages against teams with those winning percentages are 51.7% and 41.3%.*)

But, as ever, the same sizes are tiny and there is a huge amount of texture not accounted for either in brute win-x-loss or in adjusting it for strength of schedule. (E.g., I certainly didn't take home-away records into account, which is a known factor, and so on. And it's a crude way of going about things even for such an adjustment for s-o-s: one should 'really' calculate the "expected" win probability for each game and then sum those (and divide by #ofgames).)

--

* See, eg, https://sabr.org/journal/article/probabi...-matchups/

(01-01-2022, 01:55 PM)mavsluvr Wrote: [ -> ]Believe me, every time I venture into this breach, I ask myself why, lol. This time, I was fresh off reading Seth's book, and thought maybe sharing a few of his insights would elicit some interesting discussion.

Thanks for sharing! It's always good to get the perspective of someone who is not just any ol' amateur. 

(01-01-2022, 01:28 PM)fifteenth Wrote: [ -> ]I'll defend the driver of this particular conversation, even though I don't see +/- the same way he does. Kam isn't a "hater", he's been a lover of +/- for the decade I've known him. He's very consistant. He likes to use +/- and combine it with his eye test to make observations. I can't see any "hating" going on.

These comments aren't even aimed particularly at Special K, although I admit that I do think that, just every once in a while, he sometimes veers off into repeatedly citing a curated collection of +/- stats to support whatever his contrarian hobby horse du jour might be. And that can come across as hater-like (Shawn Marion, and now Luka), even if it isn't really meant that way. He's a good dawg and prolific contributor, I love him as much as the next guy, and my disagreement that the stats he cites always support his conclusions is in no way personal. But I think there are occasionally other people (maybe not so much on this board as in the Mavs media-verse at large) who do employ dodgy selections of this kind of stat in the service of hate-mongering, and that's what I was referring to. I see them on Twitter every day. Iztok had an exchange with them one day, which made me chuckle.  

I also think that you are right in your general point of view that these sorts of disagreements can provoke some excellent discussions, and what I was really hoping for was to get some points of view that might further elucidate an interesting subject.

(01-01-2022, 01:55 PM)mavsluvr Wrote: [ -> ]Believe me, every time I venture into this breach, I ask myself why, lol. This time, I was fresh off reading Seth's book, and thought maybe sharing a few of his insights would elicit some interesting discussion.

I haven't read the book, but I have now read some Q&As with him. In general terms (and not focusing on +/- (which I'm not that keen on, at least on an individual level), he comes across as trying to face both ways - as trying to take care to discuss the limitations of what basketball stats can tell one whilst advocating for use of certain of them in some, and being conciliatory with sceptics. Maybe the book takes a more robust line, but my overall impression is that probably most readers could come away thinking their prior has been confirmed by what they read - statsy people thinking 'yeah, stats are good, but need to be done carefully' and non-statsy people thinking 'yeah, there's no point in any of that really - there's so many unaccounted things behind each number' (what the statsy people might call 'noise').

(01-01-2022, 02:03 PM)BackToSquareOne Wrote: [ -> ]My point was that stats are typically hard(ish) and wins-x-losses is also one of those stats and not some touchstone that one can turn to when one is confronted with a different stat (whether or not that stat is telling one something). 

(For what it's worth the Mavs "expected" win %ages against teams with those winning percentages are 51.7% and 41.3%.*)

But, as ever, the same sizes are tiny and there is a huge amount of texture not accounted for either in brute win-x-loss or in adjusting it for strength of schedule. (E.g., I certainly didn't take home-away records into account, which is a known factor, and so on.)

--

* See, eg, https://sabr.org/journal/article/probabi...-matchups/

Thanks for the info and finding the article, BTSO. I didn't think your comments about winning percentages were hate-motivated at all, that's why I said I wasn't referring to you in particular. Agree that W-L record is subject to some of the same vagaries of chance that other stats are. Just trying to bring back the focus to the point that the goal in a season (normally) is to win games, not to accumulate the highest possible net rating (although that may actually occur as an incident of trying to win).

(01-01-2022, 02:12 PM)mavsluvr Wrote: [ -> ]Shawn Marion

Ha ha!! Were you around, maybe under a different pseudonym, when Kam and I argued about Marion for 2 or 3 years?

(01-01-2022, 02:21 PM)Tyler Wrote: [ -> ]https://twitter.com/espn_macmahon/status...1610913796

Great news. Very interested to see if this thing will ever get on track for the season.

(01-01-2022, 02:17 PM)BackToSquareOne Wrote: [ -> ]I haven't read the book, but I have now read some Q&As with him. In general terms (and not focusing on +/- (which I'm not that keen on, at least on an individual level), he comes across as trying to face both ways - as trying to take care to discuss the limitations of what basketball stats can tell one whilst advocating for use of certain of them in some, and being conciliatory with sceptics. Maybe the book takes a more robust line, but my overall impression is that probably most readers could come away thinking their prior has been confirmed by what they read - statsy people thinking 'yeah, stats are good, but need to be done carefully' and non-statsy people thinking 'yeah, there's no point in any of that really - there's so many unaccounted things behind each number' (what the statsy people might call 'noise').

Interesting. In the book, he comes across (imo) as approaching the subject as a scholar or a teacher might, rather than advancing any particular agenda. I didn't so much come away with a sense that my "priors" were confirmed or denied, as much as feeling that I had learned some stuff about the way front offices use these kinds of data points.

(01-01-2022, 02:23 PM)KillerLeft Wrote: [ -> ]Great news. Very interested to see if this thing will ever get on track for the season.

Luka and Maxi back is great news. THJ coming back and thieving minutes from DFS, JB, FN, JG, and even RB is

via GIPHY

(01-01-2022, 02:28 PM)Scott41theMavs Wrote: [ -> ]THJ coming back and thieving minutes from DFS, JB, FN, JG, and even RB is

Disagree. 

If the team can't figure out how to make THJ work in the new approach, the team's ceiling is much lower than what we hoped coming into the season. I'm not saying he's definitely going to get back on track, but I do feel like the best case scenario will only become reality if he does.

(01-01-2022, 02:21 PM)fifteenth Wrote: [ -> ]Ha ha!! Were you around, maybe under a different pseudonym, when Kam and I argued about Marion for 2 or 3 years?

I was definitely around then and posting under my current handle, which is actually the only name I have ever used on this forum or any of its predecessors. 

At the time, I was really puzzled by what seemed to me to be K's obsessive sports-hate of Marion. Now, I kind of expect it to be aimed at somebody on the team, and know that it's not necessarily hate-driven. 

Hahaha, those were the days, my friend.

---

(01-01-2022, 02:06 PM)fifteenth Wrote: [ -> ]Thanks for sharing! It's always good to get the perspective of someone who is not just any ol' amateur. 

LOL

(01-01-2022, 02:26 PM)mavsluvr Wrote: [ -> ]Interesting. In the book, he comes across (imo) as approaching the subject as a scholar or a teacher might, rather than advancing any particular agenda. I didn't so much come away with a sense that my "priors" were confirmed or denied, as much as feeling that I had learned some stuff about the way front offices use these kinds of data points.

That's interesting, I should probably put the book on my "to read" list.   I certainly can see the careful teacher angle. 
Maybe my impression of the emphases was slanted by the questions the interviewer asked.

(I suppose I also have read a number of things on how (different) FOs use various stats, so maybe I wrongly discounted that part somewhat as 'something everyone knows'.)

--

(Also, don't worry .. I didn't take any of the conversation as constituting digs at all.)

(01-01-2022, 01:56 PM)mvossman Wrote: [ -> ]Yes, this offense would be more fun to watch if they were always going against the Kings, or the Nurkless Blazers or a bunch of Covid replacement players.  Do you really think that's because there is no Luka?

What is your point with all of this anyways?  Do you really think this team is better without Luka?  Think its time to trade him for a bushel of picks?  Where are you going with this?

It wasn't the opponent that made the basketball watchable, it was the WAY they played, the way they passed, the way they worked, the way they competed. 


I don't have ANY conclusion about the causation link with Luka and how the Mavs have played this season. NONE. I will say it one more time for those that are failing at reading comprehension: I do NOT have a conclusion about whether Luka is part of the cause of the Mavs sucking when he plays this season. He might be, he might not be. Probability says he is at least a significant part, but we cannot be certain. 


What's my point with all this? To share the freaking data for discussion and rumination, but apparently that makes me some enemy and "hater." 

I have NEVER said the team is better off with Luka. I have NEVER said the Mavs should trade Luka. 


But it is a FACT that the Mavs have SUCKED at basketball overall with Luka playing this year. And that is incredibly concerning. Maybe we can glean some tentative causes carefully, maybe we cannot. But when you suck at basketball as a team when your best player plays you have MAJOR ISSUES. But I think most folks would prefer to stick their heads in the sand and pretend everything is fine with Luka. 

And I do NOT know causes in all this: But if Luka gets credit for making the team GREAT, then he should be equally culpable at getting blame when the team SUCKS. It cuts both ways. He is the superstar, he is the face of the franchise. For better or worse things start and end with him. Just as it did with Dirk.


So where am I going in all this? The Mavs have some major work to do to figure out how to be successful in outscoring their opponents when Luka plays (it is their single most important task ahead of them and it is the reason we might see major roster changes). He is going to be the one they lean on in the playoffs (if they get there), so they need to figure out winning with him on the court. They need to figure out how much of the issue is with Luka, how much is the roster, how much is the scheme, how much is coaching, how much is random noise, etc.

(01-01-2022, 02:38 PM)Kammrath Wrote: [ -> ]It wasn't the opponent that made the basketball watchable, it was the WAY they played, the way they passed, the way they worked, the way they competed. 


I don't have ANY conclusion about the causation link with Luka and how the Mavs have played this season. NONE. I will say it one more time for those that are failing at reading comprehension: I do NOT have a conclusion about whether Luka is part of the cause of the Mavs sucking when he plays this season. He might be, he might not be. Probability says he is at least a significant part, but we cannot be certain. 


What's my point with all this? To share the freaking data for discussion and rumination, but apparently that makes me some enemy and "hater." 

I have NEVER said the team is better off with Luka. I have NEVER said the Mavs should trade Luka. 


But it is a FACT that the Mavs have SUCKED at basketball overall with Luka playing this year. And that is incredibly concerning. Maybe we can glean some tentative causes carefully, maybe we cannot. But when you suck at basketball as a team when your best player plays you have MAJOR ISSUES. But I think most folks would prefer to stick their heads in the sand and pretend everything is fine with Luka. 

And I do NOT know causes in all this: But if Luka gets credit for making the team GREAT, then he should be equally culpable at getting blame when the team SUCKS. It cuts both ways. He is the superstar, he is the face of the franchise. For better or worse things start and end with him. Just as it did with Dirk.


So where am I going in all this? The Mavs have some major work to do to figure out how to be successful in outscoring their opponents when Luka plays (it is their single most important task ahead of them and it is the reason we might see major roster changes). He is going to be the one they lean on in the playoffs (if they get there), so they need to figure out winning with him on the court. They need to figure out how much of the issue is with Luka, how much is the roster, how much is the scheme, how much is coaching, how much is random noise, etc.

I think it's on the coaching and scheme. Kidd has seen a lot during Luka's absence and has adjusted well. He needs to figure out how to integrate the rest of the team with Luka and vice versa, and get Luka on board with it.

While we're exchanging views, I'll just toss in a couple more items. --

I totally disagree that the Mavs have played their best basketball of the season during the COVID interval. If I'm wrong and this is the best they can do, then I have deep despair about the Mavs. 

I am also struck by the idea that tiny sample sizes are statistically significant. One of the cornerstones of statistics is that when you use sample sizes that are too small to be representative, it can and does lead you into wildly improbably conclusions. Partnow emphasizes something we probably all viscerally know -- that misleading data (such as that supplied by mini-sample sizes) provides a significantly worse basis for reaching conclusion than no data.

(01-01-2022, 02:30 PM)KillerLeft Wrote: [ -> ]Disagree. 

If the team can't figure out how to make THJ work in the new approach, the team's ceiling is much lower than what we hoped coming into the season. I'm not saying he's definitely going to get back on track, but I do feel like the best case scenario will only become reality if he does.

We will agree to disagree emphatically, then. All of the players I listed aside from JB are far better defenders than THJ. The Mavs already have to figure out how to integrate Luka and Brunson together into Kidd's vision of defense first. THJ as well is a bridge way too far. He simply has no future here whatsoever presuming JB is re-signed and Kidd is still the coach next year. SB is also a better defender than THJ.

(01-01-2022, 02:49 PM)mavsluvr Wrote: [ -> ]I am also struck by the idea that tiny sample sizes are statistically significant. One of the cornerstones of statistics is that when you use sample sizes that are too small to be representative, it can and does lead you into wildly improbably conclusions. Partnow emphasizes something we probably all viscerally know -- that misleading data (such as that supplied by mini-sample sizes) provides a significantly worse basis for reaching conclusion than no data.

So if your wife cheated on you once you wouldn't find that "statistically significant" and would just move on with life as normal? How many times does she have to cheat on you for it to be enough data for you to pay attention to?


The point: Even a single data point is INCREDIBLY significant and is telling you something. Obviously MORE data is helpful, but oftentimes in life your data sets are small and you have to discern with wisdom how to interpret what is going on. The NFL deals with this ALL the time. Y'all are having fun mocking me for using a ten game set, but damn in the NFL that is a TON of data.

(01-01-2022, 02:50 PM)Scott41theMavs Wrote: [ -> ]All of the players I listed aside from JB are far better defenders than THJ.

Maybe, but the offense is the problem this year. The defense isn't perfect, but it's improving. This is one of the most pitiful offenses I've seen in quite some time, which is shocking given where the team was several months ago. 

I'm not even saying that making it work with THJ at any cost is necessarily the best option. But, simply taking him out of the rotation would be insane. You either have to get him back on track or use him productively in a trade. If your point is that the players you mentioned are just better overall players, I'd say that's anti-THJ bias. The version of THJ the Mavs got the previous two seasons is better than any of them, which is why he got paid more than any of them. If the Mavs can get that player going again, trust me, they'll be better off for it. If they can't because he doesn't fit here in Kidd's system as well, then they should trade him, ASAP. Simply letting others take his place in line during the first year of his new contract seems like the dumbest possible way of approaching the problem to me.