ImageImageImageImageImage

Wizards trade for Davis Bertans "the Latvian Laser"

Moderators: montestewart, nate33, LyricalRico

payitforward
RealGM
Posts: 12,233
And1: 3,822
Joined: May 02, 2012
Location: On the Atlantic

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#81 » by payitforward » Mon Aug 5, 2019 3:02 am

doclinkin wrote:
payitforward wrote:Glad to hear it, & I like Bertans. I was aware of him in the run up to the '11 draft & thought he'd be an NBA player. He's a tremendous shooter, an absolutely awful rebounder, & about average on everything else.

If you are playing an average NBA 4, then you take him out of the game & replace him with Davis Bertans, the result is that you get worse. Period.

Why, otherwise, do you think SA would want to replace him with one of the Morris twins? Why, otherwise, would you imagine that he was available to us for... nothing. Zip. Zero.


Because San Antonio has 12 players on their roster that are perimeter players. And you’re thinking of Bertans as a 4 not a 3. Here he will play that role. A face up perimeter player. A giant 2-guard. A top of the key Small Forward. Durant lite.

The Spurs have that guy four deep at all outside positions. They needed defense and extra fouls and veteran depth up front behind Aldridge considering they were losing an aging Pau. And the West is getting bigger. Well, the league is all over.

But if you look at his FIBA play, Davis is a perimeter player all the way. And if you notice, many of the players we picked up have FIBA experience. A front office puts their stamp on a team. If you were trading and drafting you would make your choice based on Rebounding first. You weight Offensive rebounds highly.

This team had opportunities to choose those players and bypassed them. Why? They have an analytics department. Tommy is a numbers guy. They talk about having a system for weighting players. Carefully choosing players based on their own internal calculus.

Now Unless you think they’re lying, or making bad choices, or simply passively selecting whatever was offered, you have to take them at their word. And if so ask yourself what is the common thread if any in each of the players they made an effort to acquire.

Our selections in the draft and in fee agency seem to be weighted towards TS%. That and ball handling are the two areas they chose to address.

Now you may not like the players they picked up. You may hope our front office is just filling slots. But you’re baffled that we didn’t make an effort to land Vonleh. Clarke. Or better rebounders with the Admiral selection.

CJ Miles, veteran perimeter forward. Bertans, perimeter forward with FIBA experience. Wagner, perimeter 4/5, played in the U19 FIBA leagues. Bryant, noted in drafting for his smooth comfortable outside stroke. Hachimura, they had their eye on him since before Gonzaga, in FIBA play he’s even more face up oriented.

Bertans is not a rebounder. He’s not a low post player at all, so you’re bound to be underwhelmed especially in your pet fetish stat of Offensive boards. In your metric he is “just a guy”.

But on this team, in year one of a deliberate rebuild, he is a starter for us. And that will definitely shape the team and how we play.

Stick your head in the sand all you want. But this team is being built around ranged gunning and efficient scoring. Outside scoring, with drive-and-kick guards.

Don’t argue with me about it if you have issue, email their analytics guy who is sending Ted 96 page long missives on the proper direction of the team.

Even Bryant, who Tommy stormed in and demanded Ernie acquired, had a strong record of efficient scoring as his primary skill in his resume. Not dominant rim defense and boardwork. Outside shooting. Catch and dunk in traffic. Good hands on the roll.

That is the team that this FO is trying to build. A team that can outgun opponents.

Unless of course there is no analysis whatsoever and they’re simply taking whatever the wind blows them as is your theory. Seems unlikely to me. But given that this team has emphasized analytics, take a look at the stats yourself and guess what sort of team you think they’re building. Make your case. Seems to me you’ll find that TS% number is the standout in the bigs they’re picking up. Look at college. Look at FIBA. See what I’m talking about. Challenge your own opinions. Guess their chess game.

FWIW you know me. I predicted teams would or ought to go big and heavy on defense and rebounding as a counter to the outside metagame. I’ve been banging on that for a year. That these players are undervalued. I wanted Vonleh mid season. I agree with your assessment on all but the value of NCAA offensive rebounds in draft assessment. This FO is working a different angle than me. Tommy May have his prejudices from all those years scouting the international game. He may be wrong.

But from everything I can see this is clearly deliberate. Theres a direction in how we are going to play this year. And with Wall as well, since they got a Wall Jr. in Ish Smith.

Actually, doc, that is an extremely well-argued point of view. & interesting as hell too. I applaud you for it. & I mean that sincerely.

...Along those same lines, I'd appreciate if for once and all you'd stop using ugly words on me ("fetish stat" in this case). Ok? I mean it, man -- don't do it again. It's not how I address you, & you are hearing me say straight out that I don't like it. We've known each other quite a while. On this board & on one preceding this one. Respect me enough to make the effort to change the way you address me....

..."I mean that sincerely." But, that doesn't necessarily mean I agree altogether. Though, in fairness, I'm seeing your point of view w/ a bit more clarity. & I absolutely agree that if Bertans is meant to play the 3 not the 4, there is a far higher likelihood of his being a positive addition.

That said, a) I haven't seen Bertans play the 3 in the NBA & don't know that he can; b) Miles is washed up, period -- he's two years away from the only really good season he ever had, & he ain't getting back there at this stage of his career; c) the issue with Wagner isn't what kind of player he is but how good a player he is, & the answer is that he isn't good at all; d) the LA trade is absolutely a case of us getting lucky, the wind happening to blow something good in our direction (& in that sense, you bet, it's just like our acquisition of Bryant); e) Mo Wagner had a low TS% not a high one (tho he had a high TS% at Michigan); f) Tommy signed McRae right off -- he doesn't shoot the 3, but you don't count him as a data point against your interpretation of Tommy's intensions; g) Bonga doesn't shoot the 3 either; h) Jemerrio Jones doesn't either i) your point about Wall/Smith doesn't work -- neither of them shoot the 3 well (or especially often -- under 2.5 times per 40 minutes on Smith's career).

As to stats -- basketball is two entirely different things. It's entertainment, & it's competition. Any number of things can make basketball entertaining. As to competition, wins/losses, they are entirely & 100% determined by numbers. Nothing else.

Obviously you can tell who won a game just by looking at the final score. & the final score is determined by two things & two things only: TS% & number of opportunities to score. Directly determined & solely determined by those 2 numbers.

Therefore the value (positive or negative) of something a player does on the floor can be given a value based on how much it affects those 2 things (positively or negatively). That's true of each & every thing a player does. Taking a shot is using up something; it has a cost. Making a shot has a positive value. Ditto rebounds, both on offense & defense. & regression analysis shows that an offensive rebound has just about twice the positive effect on outcome of a defensive rebound. Therefore, when we look at a player's rebounding we have to look at both d & o boards. A guy who gets 8 boards per 40 minutes, 2 of them offensive, helps his team win games a little more than a guy who gets 8 boards per 40 minutes, 1 of them offensive. Not much in life is simple, but that is.

Thus, no, I don't have a "fetish" about offensive boards; I just value them the regression analysis using SAS or other statistical software says they should be valued.

As to the translation from college to the pros, I don't think there's any secret sauce -- e.g. that offensive boards in college have some special value -- just that guys who have better numbers overall (valued appropriately) tend to do better as pros than those whose numbers aren't as good. Notice the phrase "tend to." After all, we're talking about kids here, they are mutable. If it was possible to predict this with a really high level of accuracy, then draft position would correspond to NBA productivity a lot more closely.

Ok... end of screed!
Remember -- if you don't like the post above: blame Doc not me.
User avatar
doclinkin
General Manager
Posts: 9,417
And1: 1,579
Joined: Jul 26, 2004
Location: .wizuds.

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#82 » by doclinkin » Mon Aug 5, 2019 4:16 am

payitforward wrote:Actually, doc, that is an extremely well-argued point of view. & interesting as hell too. I applaud you for it. & I mean that sincerely.

...Along those same lines, I'd appreciate if for once and all you'd stop using ugly words on me ("fetish stat" in this case). Ok? I mean it, man -- don't do it again. It's not how I address you, & you are hearing me say straight out that I don't like it. We've known each other quite a while. On this board & on one preceding this one. Respect me enough to make the effort to change the way you address me....



First of all. How is a tough cat from the Windy City so scorched by a stray word. I don’t take this all that seriously and don’t expect you to. It’s because I appreciate you that I’ll give you the business. I respect you enough to choose the right words to properly insult you. :clown:

And I expect to get back what I give. Or better. Since you’re the professional wordslinger. So relax Petunia. Or sling a word back at me. I like you just fine. You clearly have no problem giving hell to other people. I don’t mind if you lob it right back at me. Shoot I started the insults thread specifically so people _could_ clown on me and vice versa. I know half the board here will smirk at the irony that you of all people get insulted by a tone.

Sorry man, If you’re ever truly offended by anything I say it’s not my intent. Ought to be clear. I’m talking about your ideas, not your family or fashion choices or age or whatever. You’ve got interesting ideas even when I disagree with the height of the horse from which you pronounce them. :)

As for the rest.

..."I mean that sincerely." But, that doesn't necessarily mean I agree altogether. Though, in fairness, I'm seeing your point of view w/ a bit more clarity. & I absolutely agree that if Bertans is meant to play the 3 not the 4, there is a far higher likelihood of his being a positive addition.

That said, a) I haven't seen Bertans play the 3 in the NBA & don't know that he can; b) Miles is washed up, period -- he's two years away from the only really good season he ever had, & he ain't getting back there at this stage of his career; c) the issue with Wagner isn't what kind of player he is but how good a player he is, & the answer is that he isn't good at all; d) the LA trade is absolutely a case of us getting lucky, the wind happening to blow something good in our direction (& in that sense, you bet, it's just like our acquisition of Bryant); e) Mo Wagner had a low TS% not a high one (tho he had a high TS% at Michigan); f) Tommy signed McRae right off -- he doesn't shoot the 3, but you don't count him as a data point against your interpretation of Tommy's intensions; .


What did I say. Bigs who shoot from range. And ball handlers who can drive. But McRae is the prime example. What was he known for in his G League play this year. His defense? He went on a tear of games scoring buckets for his squad. He’s a ball handler who can drive the interior and score.

As for Wagner. As always you have a decided recency bias, but every front office has been scouting their pet players for a long time. And probably the earlier the tabbed then the more they like them later. Wagner in Michigan and prior to it as a teen was shooting well from outside. He also was touted as lighting it up in the Wizards draft workout.

Every FO wants to think their guys are going to develop for them even if they foundered elsewhere. Maybe so. Many guys struggle their first year in the league. The Wiz likely feel they’ve gotten a steal since they’ve seen him do it at the lower levels.

I notice you left off Admiral. A forward who also shot well from outside.

g) Bonga doesn't shoot the 3 either; h) Jemerrio Jones doesn't either i) your point about Wall/Smith doesn't work -- neither of them shoot the 3 well (or especially often -- under 2.5 times per 40 minutes on Smith's career)


Bonga: ballhandling face up guy. He drives the interior.

Ditto John Wall. Ish. IT. Robinson.

You’ve never drawn up plays on a napkin? Think like a coach. Put it this way, if you had John Wall on your team and wanted to build around him. Or had an owner who clearly wanted to build around him. If you had committed significant resources in contracts to him as a focal point if the team. What kind of team would you have to build to maximize his value?

Strengths: Handle. Speed. Vision. Dribble drive attack. Finding the open man, especially in the corner.

Statistically it was shown that John Wall was responsible for more corner assists for three than any other player. And those assists scored at a higher rate of success. There are any number of articles you can find on a search. But recall that the team has in game situational data on how players are scoring best next to other players.

John makes perimeter shooters look damn good.

The problem is we’ve never had anyone on the team who played like Wall when he was out of the game. So we ended up with a slower ball control attack like our Everybody Eats crew (Satoransky, Otto, Beal). Then an entirely different playset has to be input. Which worked fine until scouting caught up and realized Beal was our only threat. Nobody else wanted the ball or were able to make anything happen with it.

Now. We add ballhandling guards who drive and attack. NONE of the guards we added shoot great from outside. They all drive and pass.

As to stats --
Bla bla bla.. let me fast forward past the usual PIFfle wherein our buddy instructs everybody’s grandma how to suck eggs. Or drink berri juice — ok here:

regression analysis shows that an offensive rebound has just about twice the positive effect on outcome of a defensive rebound. Therefore, when we look at a player's rebounding we have to look at both d & o boards. A guy who gets 8 boards per 40 minutes, 2 of them offensive, helps his team win games a little more than a guy who gets 8 boards per 40 minutes, 1 of them offensive. Not much in life is simple, but that is.


Except it’s not. Because while for an individual perspective an offensive rebound looks good, from a team perspective an emphasis on offensive boards has a negative correlation with wins. It ruins court balance and leaves you out of position for transition defense.

So. If you have a particular player who is an advanced offensive boardsman okay. Let them on an island snatch those boards. While everyone else stays in their lane and plays their role. But if as a team you crash those boards. Then you tend to give up easy buckets at the other end. It’s easier for a defensive guy to set a base and box out than an athletic offensive guy to leap over him to snatch it. Why the Draymonds and Millsapps of the league succeed despite being ground bound and not the quickest.

This has changed a little bit with the increase in three point shooting. The bounces are longer. And your quicker players are closer to them. And there are more of those misses available since even the best three point teams are still missing more than ~64% of those shots. It’s a better bet that you can stall to see if you can get a stray bounce. If you have guards who can snatch em.

Guess what. We do. And we are loading up on Bigs who effectively or not will shoot from outside. Allowing our quick strong guards (Beal and Wall, and Bonga maybe) to catch the reset button. Though now you only get 14 seconds to go again. Speeding up the clock and getting more possessions per game.

To me both strategy in team building and tactics in coaching are the fun of the game. Seeing how you can assemble a squad to make those statistics work for you. It’s easy to get lost in the numbers but not think about what they mean to the coaching staff who have to deal with the four dimensional slide rule calculus of how to maximize every players effect on court. Even with overhead cameras and AI pointing out where the hot zones are and who is clicking with whom on which plays.

As to the translation from college to the pros, I don't think there's any secret sauce -- e.g. that offensive boards in college have some special value -- just that guys who have better numbers overall (valued appropriately) tend to do better as pros than those whose numbers aren't as good. Notice the phrase "tend to." After all, we're talking about kids here, they are mutable. If it was possible to predict this with a really high level of accuracy, then draft position would correspond to NBA productivity a lot more closely.

Ok... end of screed!


Check blocks and offensive boards. NCAA to the pros. The standouts on both rarely translate to the next level. Often these are guys who are taller or more athletic than their competition at the lower level, but at the next level they are out of position and jumping at everything. Foul machines.

Bigs who steal and pass well seem to defend well at the NBA level. Defensive boards seem to translate better. And D boards in non low post players seem to suggest success at the next level. Yeah it’s not perfect. But if you tease it out you can see the general shape of it. Who’s the last great NCAA offensive boardsman who translated that skill to the next level. Without looking it up I’m guessing Unibrow? Who else.
User avatar
doclinkin
General Manager
Posts: 9,417
And1: 1,579
Joined: Jul 26, 2004
Location: .wizuds.

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#83 » by doclinkin » Mon Aug 5, 2019 4:31 am

But I admit. Jones doesn’t fit the outlines of who they added. He really doesn’t look much like any other player I’ve seen. Pretty remarkable. I hope we keep him. I’m okay losing Tarik Philip for instance. Just to keep one “sic-’em” defender on the squad.
User avatar
nate33
Forum Mod - Wizards
Forum Mod - Wizards
Posts: 54,253
And1: 8,448
Joined: Oct 28, 2002

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#84 » by nate33 » Mon Aug 5, 2019 1:02 pm

doclinkin wrote:
regression analysis shows that an offensive rebound has just about twice the positive effect on outcome of a defensive rebound. Therefore, when we look at a player's rebounding we have to look at both d & o boards. A guy who gets 8 boards per 40 minutes, 2 of them offensive, helps his team win games a little more than a guy who gets 8 boards per 40 minutes, 1 of them offensive. Not much in life is simple, but that is.


Except it’s not. Because while for an individual perspective an offensive rebound looks good, from a team perspective an emphasis on offensive boards has a negative correlation with wins. It ruins court balance and leaves you out of position for transition defense.

So. If you have a particular player who is an advanced offensive boardsman okay. Let them on an island snatch those boards. While everyone else stays in their lane and plays their role. But if as a team you crash those boards. Then you tend to give up easy buckets at the other end. It’s easier for a defensive guy to set a base and box out than an athletic offensive guy to leap over him to snatch it. Why the Draymonds and Millsapps of the league succeed despite being ground bound and not the quickest.

This is an interesting discussion, and since we don't have much else to talk about this offseason, I think it's worth exploring further.

Looking at the team statistics for last season, it would appear that Doc's argument about the negative correlation between offensive boards and wins holds water. Here are the 30 teams, ranked by offensive rebound percentage.

Image

Pay attention to the NRtg column in the middle which lets you know how teams fared. With the exception of the 3 teams at the top, there is a whole lot of red ink among the 12 teams with an above average ORb%. And two of those top three teams that utilized offensive rebounding as a successful strategy were OKC and Portland, teams with exceptional one-on-one ball handlers who can create without the benefit of much spacing. Basically, they give the ball to Lillard, McCollum, Westbrook or George, tell them to jack up a shot, and tell everyone else to crash the glass. That only works if you have elite talent handling the ball.

With the exception of Denver, and the teams with elite iso scorers, offensive rebounding did not translate to wins. The other two modestly good offensive rebounding teams: Philly and Utah, mostly got their offensive rebounds by accident, by virtue of having a behemoth at center. It wasn't necessarily a strategy.

Meanwhile, most of the competent playoff teams including 5 of the 6 playoff teams in the East with a positive point differential, had a substantially below-average offensive rebound rate. The best team in the league, with a pair of 7-footers in the starting lineup, was the 4th worst offensive rebounding team.
User avatar
nate33
Forum Mod - Wizards
Forum Mod - Wizards
Posts: 54,253
And1: 8,448
Joined: Oct 28, 2002

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#85 » by nate33 » Mon Aug 5, 2019 1:05 pm

It appears that defensive rebounding has a much higher correlation with success:

Image

The Rockets and Pacers are the only good teams with a pretty bad defensive rebounding percentage. And literally all of the top 10 defensive rebounding teams made the playoffs.
payitforward
RealGM
Posts: 12,233
And1: 3,822
Joined: May 02, 2012
Location: On the Atlantic

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#86 » by payitforward » Mon Aug 5, 2019 1:58 pm

doclinkin wrote:But I admit. Jones doesn’t fit the outlines of who they added. He really doesn’t look much like any other player I’ve seen. Pretty remarkable. I hope we keep him. I’m okay losing Tarik Philip for instance. Just to keep one “sic-’em” defender on the squad.

Fine. You've conceded my single most critical point. Jemerrio Jones is a must keep. Alas, to keep him we have to let 2 guys go -- Phillip & one other, who I think would have to be McRae (i.e. I keep Robinson over McRae).
Remember -- if you don't like the post above: blame Doc not me.
payitforward
RealGM
Posts: 12,233
And1: 3,822
Joined: May 02, 2012
Location: On the Atlantic

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#87 » by payitforward » Mon Aug 5, 2019 3:16 pm

nate33 wrote:
doclinkin wrote:
payitforward wrote:regression analysis shows that an offensive rebound has just about twice the positive effect on outcome of a defensive rebound....

Except it’s not. Because while for an individual perspective an offensive rebound looks good, from a team perspective an emphasis on offensive boards has a negative correlation with wins...

This is an interesting discussion, and since we don't have much else to talk about this offseason, I think it's worth exploring further.

Looking at the team statistics for last season, it would appear that Doc's argument about the negative correlation between offensive boards and wins holds water.

Sorry, no. But, even if Doc were correct it would be irrelevant to my point, because I didn't say the thing he's attempting to correct. However, he isn't right. Lets look at your chart, nate.

nate33 wrote:Here are the 30 teams, ranked by offensive rebound percentage.

Image

First off, this chart is utterly unrelated to the point I make about the value of an offensive rebound. You will search in vain for a place where I say the best teams get the highest % of offensive rebounds. Establishing that the value of an offensive rebound @2 times that of a defensive rebound is unrelated to whether a team should hunt for offensive rebounds.

I don't have the patience to go through this again, but, I think I can make it clear in a comparison to another field of activity. What is the value of a stock that triples in six months? That's easy, duh.

So... should one concentrate one's investing strategy exclusively or even just predominantly on finding stocks that have a chance to triple in six months? No, b/c if you do so you'll go broke! Duh.

Does that make a stock that triples in six months any less valuable? No.

Get my point? Yet, nate, in any case your chart doesn't say what you think it says.
nate33 wrote:Pay attention to the NRtg column in the middle which lets you know how teams fared. With the exception of the 3 teams at the top, there is a whole lot of red ink among the 12 teams with an above average ORb%....

Absolutely, lets give that column some attention. Only... you can't just look at the distribution of red ink; you have to look at the actual numbers. For example:

1. The top 5 teams in ORB% overall are net +11.1 in NRtg overall. The bottom 5 teams in ORB% are -10 in NRtg overall.
2. The top 10 teams in ORB% are also better overall in NRtg than the bottom 10.
3. For utter ineradicable simplicity, compare the top 15 & the bottom 15 in ORB%. The top 15 teams in ORB% have a higher NRtg, overall than the bottom 15.

QED.

Does this mean you should hunt offensive boards as a strategy? No, of course not! Any more than the fact that a made three is worth 1.5 times a made two means that you should shoot more threes, independent of whether you have anyone on your team who is skilled at shooting them. Go back to that bit about stocks that triple.
Remember -- if you don't like the post above: blame Doc not me.
User avatar
nate33
Forum Mod - Wizards
Forum Mod - Wizards
Posts: 54,253
And1: 8,448
Joined: Oct 28, 2002

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#88 » by nate33 » Mon Aug 5, 2019 4:04 pm

payitforward wrote:First off, this chart is utterly unrelated to the point I make about the value of an offensive rebound. You will search in vain for a place where I say the best teams get the highest % of offensive rebounds. Establishing that the value of an offensive rebound @2 times that of a defensive rebound is unrelated to whether a team should hunt for offensive rebounds.

I can't believe you can't grasp this point. Yes, an individual offensive rebound is a good thing, but clearly, there are hidden costs to obtaining offensive rebounds because teams that scheme to get them don't seem to boost their win total at all.

This ties in with the argument you and I have all the time. You say a guy like Sam Dekker is the bees knees because he rebounds so well, and then you demonstrate using the same tired possessions gained/lost breakdown to try and prove your case mathematically. But clearly, based on the team performance data, this analysis isn't telling the whole story. If it did, it would be pretty easy to field a bunch of "undervalued" players who rebound really well and shoot poorly, throw them together on a team, and expect to win. But when you do so in real life, you find that spacing problems of a poor shooting team tend to have negative consequences not anticipated by your individual player calculations. Suddenly, all those great offesnive rebounders aren't rebounding so well because the other team is packing the paint. Meanwhile, the offensive rebounding team is getting fastbreaked to death at the other end of the floor.


payitforward wrote:Yet, nate, in any case your chart doesn't say what you think it says.
nate33 wrote:Pay attention to the NRtg column in the middle which lets you know how teams fared. With the exception of the 3 teams at the top, there is a whole lot of red ink among the 12 teams with an above average ORb%....

Absolutely, lets give that column some attention. Only... you can't just look at the distribution of red ink; you have to look at the actual numbers. For example:

1. The top 5 teams in ORB% overall are net +11.1 in NRtg overall. The bottom 5 teams in ORB% are -10 in NRtg overall.
2. The top 10 teams in ORB% are also better overall in NRtg than the bottom 10.
3. For utter ineradicable simplicity, compare the top 15 & the bottom 15 in ORB%. The top 15 teams in ORB% have a higher NRtg, overall than the bottom 15.

True, but the difference between the top 15 and the bottom 15 is negligible. There is effectively no correlation between team ORB% and winning, or an extraordinarily weak one at best. The top 15 teams at a skill you value so highly are no better than the worst 15 teams at that skill. Doesn't that tell you something about that skill? If it's not valuable at a team level (or rather, obtaining offensive rebounds at a team level tend to cost you in other facets of the game), then surely those dynamics apply at an individual level.
User avatar
gtn130
Analyst
Posts: 3,289
And1: 2,498
Joined: Mar 18, 2009

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#89 » by gtn130 » Mon Aug 5, 2019 4:47 pm

nate33 wrote:Yes, an individual offensive rebound is a good thing, but clearly, there are hidden costs to obtaining offensive rebounds because teams that scheme to get them don't seem to boost their win total at all.


This pretty much sums it up. Since about ~2016 the vast majority of NBA teams have been totally foregoing offensive rebounds for better transition defense. Teams figured out what Nate is saying above - there are hidden costs to obtaining offensive boards, primarily that failed attempts at offensive rebounds leave you way out of position defensively. I believe the Spurs were the first team to really start doing this.

I do think because of the current NBA meta there could be opportunities to go for offensive rebounds, but they're pretty situational and context-dependent.
User avatar
doclinkin
General Manager
Posts: 9,417
And1: 1,579
Joined: Jul 26, 2004
Location: .wizuds.

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#90 » by doclinkin » Mon Aug 5, 2019 4:50 pm

nate33 wrote:It appears that defensive rebounding has a much higher correlation with success:

Image

The Rockets and Pacers are the only good teams with a pretty bad defensive rebounding percentage. And literally all of the top 10 defensive rebounding teams made the playoffs.


Right something I've been howling about for a while. I mean it makes sense that teams with high raw Offensive board totals are not winning teams: they miss a lot of shots. But even emphasis on O boards can be risky. Except I do think it works better If you have second line rebounders like Westbrook who can snatch long bounces off of three point shots. And know when to do so. Stef Curry is a master of the re-set three.

Basically players who individually board well on offense are good. Teams that do so. Bad.

Unsurprisingly one of the strongest correlations between wins and losses is .... eFG% Shooting. In fact of the top ten best shooting teams last year, the only one that didn't make the playoffs was: your Washington Wizards.

And why we didn't succeed was quite often in that rebounding differential. And opponent scoring. Which is why it was a puzzler for me that we did nothing to address that shortfall in the players we chased.
User avatar
doclinkin
General Manager
Posts: 9,417
And1: 1,579
Joined: Jul 26, 2004
Location: .wizuds.

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#91 » by doclinkin » Mon Aug 5, 2019 4:55 pm

payitforward wrote:
doclinkin wrote:But I admit. Jones doesn’t fit the outlines of who they added. He really doesn’t look much like any other player I’ve seen. Pretty remarkable. I hope we keep him. I’m okay losing Tarik Philip for instance. Just to keep one “sic-’em” defender on the squad.

Fine. You've conceded my single most critical point. Jemerrio Jones is a must keep. Alas, to keep him we have to let 2 guys go -- Phillip & one other, who I think would have to be McRae (i.e. I keep Robinson over McRae).



HAH! You're not slick. :) Neat tactic. Find the small point of agreement then declare victory and stalk off holding your hands above your head in victory. I already said Jones was the one who was an outlier. And he was the one you were most nervous we would lose. Yeah, the last guy in a throw in trade is the one that shows their blueprint. The one who has played the fewest minutes. True. Yes that was the pivot point of your entire argument: that the team has no plan and they are just recruiting guys to fill slots. Because: Jemarrio!

You pegged it. Smh. Goofy.

That said, how many two-way contracts can we carry? Is Robinson one? I'm losing track. I know outside shooting Garrison Mathews is one.
User avatar
nate33
Forum Mod - Wizards
Forum Mod - Wizards
Posts: 54,253
And1: 8,448
Joined: Oct 28, 2002

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#92 » by nate33 » Mon Aug 5, 2019 5:27 pm

doclinkin wrote:
payitforward wrote:
doclinkin wrote:But I admit. Jones doesn’t fit the outlines of who they added. He really doesn’t look much like any other player I’ve seen. Pretty remarkable. I hope we keep him. I’m okay losing Tarik Philip for instance. Just to keep one “sic-’em” defender on the squad.

Fine. You've conceded my single most critical point. Jemerrio Jones is a must keep. Alas, to keep him we have to let 2 guys go -- Phillip & one other, who I think would have to be McRae (i.e. I keep Robinson over McRae).



HAH! You're not slick. :) Neat tactic. Find the small point of agreement then declare victory and stalk off holding your hands above your head in victory. I already said Jones was the one who was an outlier. And he was the one you were most nervous we would lose. Yeah, the last guy in a throw in trade is the one that shows their blueprint. The one who has played the fewest minutes. True. Yes that was the pivot point of your entire argument: that the team has no plan and they are just recruiting guys to fill slots. Because: Jemarrio!

You pegged it. Smh. Goofy.

That said, how many two-way contracts can we carry? Is Robinson one? I'm losing track. I know outside shooting Garrison Mathews is one.

We can carry two. Mathews is the only one that is currently locked in as a two-way player.
User avatar
nate33
Forum Mod - Wizards
Forum Mod - Wizards
Posts: 54,253
And1: 8,448
Joined: Oct 28, 2002

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#93 » by nate33 » Mon Aug 5, 2019 5:32 pm

doclinkin wrote:Right something I've been howling about for a while. I mean it makes sense that teams with high raw Offensive board totals are not winning teams: they miss a lot of shots.

Just for clarification, my chart was DRB%, not total rebounds. Rebounding percentage is not dependent on team FG%. It's just as easy to be a good (or bad) rebounding team if you shoot a high percentage as it is if you shoot a low percentage.
User avatar
doclinkin
General Manager
Posts: 9,417
And1: 1,579
Joined: Jul 26, 2004
Location: .wizuds.

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#94 » by doclinkin » Mon Aug 5, 2019 5:35 pm

nate33 wrote:
doclinkin wrote:
payitforward wrote:Fine. You've conceded my single most critical point. Jemerrio Jones is a must keep. Alas, to keep him we have to let 2 guys go -- Phillip & one other, who I think would have to be McRae (i.e. I keep Robinson over McRae).



HAH! You're not slick. :) Neat tactic. Find the small point of agreement then declare victory and stalk off holding your hands above your head in victory. I already said Jones was the one who was an outlier. And he was the one you were most nervous we would lose. Yeah, the last guy in a throw in trade is the one that shows their blueprint. The one who has played the fewest minutes. True. Yes that was the pivot point of your entire argument: that the team has no plan and they are just recruiting guys to fill slots. Because: Jemarrio!

You pegged it. Smh. Goofy.

That said, how many two-way contracts can we carry? Is Robinson one? I'm losing track. I know outside shooting Garrison Mathews is one.

We can carry two. Mathews is the only one that is currently locked in as a two-way player.


Hm. Based on Brooks' quotes it seemed Justin Robinson was a 2 way candidate. But Brooks may not have known the contract details etc.

https://www.bulletsforever.com/2019/6/21/18700421/2019-nba-draft-washington-wizards-justin-robinson-virginia-tech-undrafted-prospect
User avatar
nate33
Forum Mod - Wizards
Forum Mod - Wizards
Posts: 54,253
And1: 8,448
Joined: Oct 28, 2002

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#95 » by nate33 » Mon Aug 5, 2019 5:47 pm

doclinkin wrote:
nate33 wrote:
doclinkin wrote:

HAH! You're not slick. :) Neat tactic. Find the small point of agreement then declare victory and stalk off holding your hands above your head in victory. I already said Jones was the one who was an outlier. And he was the one you were most nervous we would lose. Yeah, the last guy in a throw in trade is the one that shows their blueprint. The one who has played the fewest minutes. True. Yes that was the pivot point of your entire argument: that the team has no plan and they are just recruiting guys to fill slots. Because: Jemarrio!

You pegged it. Smh. Goofy.

That said, how many two-way contracts can we carry? Is Robinson one? I'm losing track. I know outside shooting Garrison Mathews is one.

We can carry two. Mathews is the only one that is currently locked in as a two-way player.


Hm. Based on Brooks' quotes it seemed Justin Robinson was a 2 way candidate. But Brooks may not have known the contract details etc.

https://www.bulletsforever.com/2019/6/21/18700421/2019-nba-draft-washington-wizards-justin-robinson-virginia-tech-undrafted-prospect

According to Sportrac, Robinson got a legit vet minimum contract, not a two-way deal. And it doesn't specify that it was an Exhibit 10 contract (which facilitates transforming the contract into a two-way deal at the behest of the team).

That doesn't mean he won't play in the G-League, but it looks like he'll be taking up a roster spot while doing so.

Assuming Sportrac is accurate, then what PIF has been saying rings true. We have 18 guys on the roster at the moment, plus Mathews as a two-way guy. Phil Booth is one of those 18 but he has an Exhibit 10 contract so chances are, he'll become our second two-way guy, so that leaves 17. (He could also be cut.)

We need to get it down to 15. Of the 17 players, the only guys with non-guaranteed deals are Tarik Phillip, Jordan McRae and Jemerrio Jones. Robinson has only $250,000 guaranteed. Two of them must go. Maybe whomever is cut can be resigned on a two-way deal (if we end up cutting Phil Booth), but all but Robinson are pretty old and may opt to play overseas for bigger money rather than continue to toil in the G-League for peanuts.
payitforward
RealGM
Posts: 12,233
And1: 3,822
Joined: May 02, 2012
Location: On the Atlantic

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#96 » by payitforward » Mon Aug 5, 2019 7:54 pm

2 different matters raised in your post; I'll take them up separately. Actually, more than one in this half for that matter!
nate33 wrote:
payitforward wrote:First off, this chart is utterly unrelated to the point I make about the value of an offensive rebound. You will search in vain for a place where I say the best teams get the highest % of offensive rebounds. Establishing that the value of an offensive rebound @2 times that of a defensive rebound is unrelated to whether a team should hunt for offensive rebounds.

I can't believe you can't grasp this point. Yes, an individual offensive rebound is a good thing, but clearly, there are hidden costs to obtaining offensive rebounds because teams that scheme to get them don't seem to boost their win total at all....

Right back at you on the "I can't believe you can't grasp this point," nate.

I thought I'd addressed this with my narrative about stocks that triple in value, but I guess not. There are costs to scheming to get offensive boards -- just as there are costs to deciding your team should shoot a lot of threes even if you don't have guys who shoot it well. They're not "hidden" costs -- they're right there out in the open.

IOW, I *do* grasp your point. & I don't even disagree with it! Let me try a different tack. The way you establish the value of something someone does on the court (positive value or negative value) is by running statistical regressions using something like SAS (the leading statistical software program as you prob. know). That's been done a zillion times. The results say that when a single offensive board happens it contributes almost 2 times as much to a win than a single defensive board.

Now... am I going on faith that the academic sports economists who do this stuff know what they're doing? Sure I am! I'm not going to go out & do it myself! For one thing, SAS is expensive; for another I don't know how to use it; for a 3d I depend on experts in many fields all the time, as all of us do, so there's nothing weird about it.

Given that, what's the meaning? What should you do about this if you "know" it? That's where the analogy of a stock that triples came in, or I thought it would. I.e. the meaning of this fact about offensive boards, the message it sends, is most definitely *not* that you should go out & do all you can to get more offensive rebounds. Of course not! No more than you should turn yourself into a speculative investor, because *if* you succeed you get a higher multiple on your investment.

But that doesn't mean that this fact, the value of an offensive board vs. a defensive board, has *no* meaning! Lets try a thought experiment. Lets suppose I'm trying to decide between two potential FAs -- which one should I sign. & lets assume that the guys are, amazingly, exactly the same in every way across the board. They score the same # of points per 40 minutes at the same TS%, steal the ball, turn the ball over, get assists, commit fouls, etc. all exactly at the same levels. They even get the same number of rebounds per 40 minutes -- 8, lets say. & you have reason to be confident that these numbers can be relied on overall -- at least to an equal degree in the case of each of them.

If one of those guys gets 6 defensive boards & 2 offensive boards, while the other gets 7 defensive boards & 1 offensive board, there you have a data point in favor of one guy over the other, in favor of the guy who gets 6 & 2 over the guy who gets 7 & 1.

Obviously, things are never that cut & dried. So in practice, what this means is that in considering/comparing any 2 guys (or just evaluating a single player for that matter) you have to look at their offensive rebounding a little differently than their defensive rebounding.

That's it. That's all.

nate33 wrote:This ties in with the argument you and I have all the time. You say a guy like Sam Dekker is the bees knees because he rebounds so well, and then you demonstrate using the same tired possessions gained/lost breakdown to try and prove your case mathematically. ...

Nah.... First off I never said Sam Dekker was the bees knees (though I appreciate the turn of phrase!). I said he was a bargain when you considered his level of productivity across the board compared to what he'd be paid. In fact, his biggest problem is that he has mostly had to play the 4, and he does not rebound well enough for that spot!

What you do have right about me in this instance, however, is that I'd decide about Sam in just exactly this way! IOW, I'd look at his numbers per 40 minutes in all these categories, I'd compare them to other guys, & I'd come to some conclusions based on the comparison -- conclusions in which I'd have the level of confidence I had in the numbers themselves of the guys I was comparing.

I don't see what could be considered outlandish in that, nate. You didn't hesitate to quote numbers in support of Marquese Kriss being a crappy player -- & I thought you were altogether justified in doing so.

nate33 wrote:... clearly, based on the team performance data, this analysis isn't telling the whole story. If it did, it would be pretty easy to field a bunch of "undervalued" players who rebound really well and shoot poorly, throw them together on a team, and expect to win. But when you do so in real life, you find that spacing problems of a poor shooting team tend to have negative consequences not anticipated by your individual player calculations. Suddenly, all those great offesnive rebounders aren't rebounding so well because the other team is packing the paint. Meanwhile, the offensive rebounding team is getting fastbreaked to death at the other end of the floor....

But... but... but... :) I would never make this claim!! Moreover, you describe the situation straight through to its conclusion -- so if I did make the claim, I'd have to claim that they were winning in spite of the fact that the premise is that they aren't winning!

Nor do I discount "spacing" in particular -- no matter how many times anyone says I do, I still don't! Honestly. I just don't.

Three things -- 1) individual numbers can't tell you absolutely everything; in any complex situation there are other factors relating to the interaction of the individual entities that must be considered (viz. spacing as a single example); 2) those interactions are difficult to capture quantitatively with great accurac, certainly they can't be captured as accurately as the individual numbers can, which means that there is a limit to the utility of the individual numbers -- which can be expressed most simply/elegantly by way of a level of confidence in the meaning of those individual numbers -- a level which is never 100%; & 3) all the same those numbers remain the single most useful set of predictors of performance that we have.

Every one of those 3 points is a truism of statistical analysis. No one in his right mind would deny them. You can decide for yourself whether I'm in my right mind, but for sure I don't deny any of them!
Remember -- if you don't like the post above: blame Doc not me.
Ruzious
Forum Mod
Forum Mod
Posts: 38,248
And1: 5,383
Joined: Jul 17, 2001
       

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#97 » by Ruzious » Mon Aug 5, 2019 8:14 pm

Pif, you say you don't discount spacing and then say it should be ignored unless/until it can be quantified, right? I'm trying to help you out, because this has become a sticking point 275 times or so. :)
"Look, you never know when you may need to borrow a cup of sugar, maybe some milk or a handgun" - Dan C. from Texas
payitforward
RealGM
Posts: 12,233
And1: 3,822
Joined: May 02, 2012
Location: On the Atlantic

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#98 » by payitforward » Mon Aug 5, 2019 8:27 pm

nate33 wrote:
payitforward wrote:...3. For utter ineradicable simplicity, compare the top 15 & the bottom 15 in ORB%. The top 15 teams in ORB% have a higher NRtg, overall than the bottom 15.

True, but the difference between the top 15 and the bottom 15 is negligible. There is effectively no correlation between team ORB% and winning, or an extraordinarily weak one at best. The top 15 teams at a skill you value so highly are no better than the worst 15 teams at that skill. Doesn't that tell you something about that skill? If it's not valuable at a team level (or rather, obtaining offensive rebounds at a team level tend to cost you in other facets of the game), then surely those dynamics apply at an individual level.

Sigh....

The original claim -- to which I was responding -- was that a high ORB% correlated with a poor record. The more you did it the worst you were (overall, statistically...). & that this chart "proved" it. Doc's still interpreting the chart that way.

But what the chart shows is that ORB% cannot be correlated negatively with wins/losses. It shows, in other words, the exact opposite of what was claimed for it. Indeed, if there is any correlation (weak), it is with winning not losing.

But, never ever did I nor would I claim that there was such a correlation! In fact, I was shocked that there was even a weak positive correlation. For one thing, offensive boarding performance, either expressed as a % or a raw quantity, is a single little stat among a zillion others. Or, lets just say 20 others of equal significance in one way or another. & teams differ in all of them, of course.

For that reason, even if offensive rebounding % were enormously significant to NRtg (it isn't) it could never jump out in a raw results chart of this kind.

How could it? The differences of all 30 teams in every one of all 20 (putative) of those variables are represented in the NRtg of each and every team. This is NOT how you do statistical analysis, sorry. You do it by running regressions that allow you to take each variable out of the result, 1 by 1, then 2 by 2, then... etc. until you are able to actually assess (with some level of confidence not 100%) what effect any single variable is having.

If that sentence used unfamiliar term & made no sense to you, then like 100% of humans you were not born in a statistics class. Read the wikipedia article on "statistical regression."

In short, there is no useful information of any kind to be gotten from this chart on the effect of ORB% on... anything. Denying that statement would only indicate that a person has never studied statistics in any context -- not in high school & not in college. Which is also not a disability! In fact... consider yourself lucky.
Remember -- if you don't like the post above: blame Doc not me.
User avatar
gtn130
Analyst
Posts: 3,289
And1: 2,498
Joined: Mar 18, 2009

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#99 » by gtn130 » Mon Aug 5, 2019 8:27 pm

PIF if you grasp Nate's point and do not disagree, then the argument you're making is just pedantic and not useful. Your arguments consistently boil down to "xyz has an inherent undeniable value because it exists statistically" and this argument serves no real purpose on a message board for speculation about basketball.

The best player in the league is Jordan Sibert, PIF. He had a 150% eFG in 2019 by taking one single three pointer and making it. You cannot argue against this fact. He is tied as the best shooter in NBA history. It's an indisputable fact, and if you argue against this then you are simply a fool. If anyone tries to say that LeBron, Giannis or anyone else is better than him, they are ignorant of the facts.
payitforward
RealGM
Posts: 12,233
And1: 3,822
Joined: May 02, 2012
Location: On the Atlantic

Re: Wizards trade for Davis Bertans "the Latvian Laser" 

Post#100 » by payitforward » Mon Aug 5, 2019 8:44 pm

Ruzious wrote:Pif, you say you don't discount spacing and then say it should be ignored unless/until it can be quantified, right? I'm trying to help you out, because this has become a sticking point 275 times or so. :)

Thanks for asking.

I don't discount spacing. Start with this: it's obvious that it is at least conceivable that spacing would have an effect. That seems undeniable to me -- & that would be enough all on its own that one could absolutely NOT discount spacing.

If it's conceivable that something could have an effect, then it should be studied to discover whether it actually has an effect, how & to what degree & on what. In a business like sports, where there's lots of money at stake in highly competitive situations, I would assume that if something should be studied (& assuming that it can be studied, which in this case I certainly assume) it is being studied.

Studied by smart people. Who conclude that it does have an effect. That's good enough for me. Way good enough! I wouldn't for a minute question that spacing has an effect, hence no way could I discount it.

What I don't know, & it's impossible for me to figure out (or you, nate, etc.) is how to quantify that effect. So, whether we're talking about Bertans or we're talking about Ben Simmons, it's hard to bring that effect of spacing into your picture of the player's value -- either as a negative factor or a positive one.

One thing you could do is look at Davis Bertans' teammates -- guys who were with SA before he arrived & are, or will be, there now that he's gone. You could look at all their performances when he was there & when he's not. You could do the same for guys who played with him on the court & when he was off the court.

But people aren't machines, so there's a lot of variation in every one of those guys' baselines -- independent of Bertans. So it's a damned complicated thing to figure out.

Especially since it's not an independent variable exactly, but rather more of a limiting factor or an enabling one (depending on whether & to what degree it's absent or present).

Make sense?
Remember -- if you don't like the post above: blame Doc not me.

Return to Washington Wizards