COACHING 101: PLAY YOUR BEST 5

There have been a few recurring themes for me during this NBA season. The first is that I ruthlessly attack anyone that promotes the meme that "small ball" is the optimal strategy in the modern game. The optimal strategy is to play your best 5. For the Heat, this means playing LBJ at the 4 so they can get Ray Allen/Shane Battier on the floor. For the Thunder, this sometimes means Durant at the 4 so they can get Martin on the floor more (and Perkins OFF the floor more). And this works because those teams have insanely talented 3s with the physical gifts to play the 4.

For the 2007 Warriors, playing small meant that Matt Barnes and JRich got more time, and it wasn't a big sacrifice because they had a terrible front court. Johnson, of course, decided to counter with his own small ball team, which meant his best rebounder (Erick Dampier) sat on the bench. Unlike Nelson, Avery was not playing his best 5, and unsurprisingly, that didn't turn out well for his team.

Another recurring theme for me so far in the 2012-13 season is that I am pretty disappointed about the injury-ridden and underperforming Timberwolves, but at the end of the day, at least Corey Brewer is no longer a T-wolf, and George Karl is not coaching them. The Nuggets have several problems right now, and those are two of their big ones.

Here's a quick look at the Nuggets:

NAME POS GP MIN WP48 PoP48 WINS PTS REB AST TO BLK STL PF
Kenneth Faried PF 16 496 .293 6.0 3.02 20.5 17.1 0.8 2.2 1.7 1.5 5.0
JaVale McGee C 16 302 .166 2.1 1.04 26.5 14.6 1.3 3.3 4.3 0.5 4.6
Kosta Koufos C 16 351 .150 1.6 1.10 13.4 12.7 1.2 1.5 3.7 1.0 7.0
Andre Iguodala SG 16 567 .150 1.6 1.77 21.2 8.0 4.8 3.9 1.1 1.5 1.9
Jordan Hamilton SF 10 112 .127 0.9 0.30 24.4 9.4 3.0 2.6 0.9 1.3 3.4
Andre Miller PG 16 398 .112 0.4 0.93 16.0 5.4 9.9 4.6 0.2 1.4 3.5
Quincy Miller SF 0 0 .000 0.0 0.00 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Julyan Stone GF 0 0 .000 0.0 0.00 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Timofey Mozgov C 6 60 .079 -0.6 0.10 12.8 12.8 0.8 3.2 3.2 0.8 12.8
Ty Lawson PG 16 585 .053 -1.4 0.65 17.7 3.4 9.9 4.1 0.1 2.8 2.1
Anthony Randolph PF 3 18 .053 -1.4 0.02 5.3 8.0 8.0 2.7 2.7 0.0 5.3
Danilo Gallinari SF 15 507 .025 -2.3 0.26 22.0 8.8 3.3 2.2 0.9 0.8 3.3
Corey Brewer SF 16 370 .015 -2.6 0.11 21.0 6.4 1.6 2.2 0.6 2.5 3.8
Evan Fournier SG 6 54 -.037 -4.2 -0.04 21.3 2.7 5.3 6.2 0.0 0.0 5.3
Wilson Chandler GF 4 69 -.231 -10.3 -0.33 18.1 13.9 0.0 3.5 0.7 0.0 7.0

Faried is by far the best Nugget so far (an opinion which should not surprise regular readers of this blog). Yes, Faried has defensive limitations. But his net rebounds, his low turnover rate, and his shooting efficiency more than make up for this, in particular his offensive rebounding, which is behind only Anderson Varejao and Jordan Hill. He is followed by the two big men McGee and Koufos, which again, for readers of this blog, should not be surprising.

If you sort by minutes, we see that the rank changes a bit:

NAME POS GP MIN WP48 PoP48 WINS PTS REB AST TO BLK STL PF
Ty Lawson PG 16 585 .053 -1.4 0.65 17.7 3.4 9.9 4.1 0.1 2.8 2.1
Andre Iguodala SG 16 567 .150 1.6 1.77 21.2 8.0 4.8 3.9 1.1 1.5 1.9
Danilo Gallinari SF 15 507 .025 -2.3 0.26 22.0 8.8 3.3 2.2 0.9 0.8 3.3
Kenneth Faried PF 16 496 .293 6.0 3.02 20.5 17.1 0.8 2.2 1.7 1.5 5.0
Andre Miller PG 16 398 .112 0.4 0.93 16.0 5.4 9.9 4.6 0.2 1.4 3.5
Corey Brewer SF 16 370 .015 -2.6 0.11 21.0 6.4 1.6 2.2 0.6 2.5 3.8
Kosta Koufos C 16 351 .150 1.6 1.10 13.4 12.7 1.2 1.5 3.7 1.0 7.0
JaVale McGee C 16 302 .166 2.1 1.04 26.5 14.6 1.3 3.3 4.3 0.5 4.6
Jordan Hamilton SF 10 112 .127 0.9 0.30 24.4 9.4 3.0 2.6 0.9 1.3 3.4
Wilson Chandler GF 4 69 -.231 -10.3 -0.33 18.1 13.9 0.0 3.5 0.7 0.0 7.0
Timofey Mozgov C 6 60 .079 -0.6 0.10 12.8 12.8 0.8 3.2 3.2 0.8 12.8
Evan Fournier SG 6 54 -.037 -4.2 -0.04 21.3 2.7 5.3 6.2 0.0 0.0 5.3
Anthony Randolph PF 3 18 .053 -1.4 0.02 5.3 8.0 8.0 2.7 2.7 0.0 5.3
Julyan Stone GF 0 0 .000 0.0 0.00 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Quincy Miller SF 0 0 .000 0.0 0.00 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Clearly, the best players have not been getting the most minutes. Now, not all of this is truly something you can blame George Karl for. Ty Lawson and Danilo Gallinari have historically been good players, and Iguodala is usually even better than the above-average pace he's been on this year. It's not really surprising or controversial that those players would be in the top 3 in minutes (although, frankly, Lawson has earned the benefit of the doubt in prior seasons, but Gallinari has not. We'll get back to that). And it is very bad timing that all three of them are underperforming their career averages at once.

But what is surprising is that no center is in the top 6 in minutes played, and only one power forward is in the top ten. Yes, Denver, your coach thinks that Corey Brewer is more important to winning than either Kosta Koufos or JaVale McGee, and that Gallo, who in 4+ seasons has never once posted average numbers in rebounds or blocks, is the team's best power forward (and sometimes center, based on lineups I have seen!) after Kenneth Faried.

This is madness. This is not brilliant strategy, clever coaching juggling, experience, savvy, or anything like this. This is just sheer stubbornness. George Karl is just essentially being Johnny the Gamer.

What the hell does that mean? Wll, I'm glad you ask. This will make you realize why "geek" is in this blog's name. If you have ever talked to any game designers, they will often tell you that there are three kinds of gamers: Jimmy, Johnny and Spike. Think of these like customer profiles, except made by game designers to evaluate how their games will appeal to gamers. The idea originated with Magic: The Gathering but has also been applied to plenty of computer games like Diablo, World of Warcraft, Everquest, etc.

Timmy:

Timmy is what we in R&D call the "power gamer." Timmy likes to win big. He doesn’t want to eke out a last minute victory. Timmy wants to smash his opponents.

One of the misconceptions is that Timmy has to be young. While its true that younger players are more apt to fall into this category, players of any age can be a Timmy. What sets Timmy apart from the other two profiles is that Timmy is motivated by fun. He plays games because they are enjoyable. Timmy is very social. An important part of the game is sitting around with his friends.

Timmy cares more about the quality of a win than the quantity of his wins. For example, Timmy sits down and plays ten games. He only wins three games out of ten but the three he wins, he dominates his opponent. Timmy had fun. Timmy walks away happy.

Spike:

Spike is the competitive player. Spike plays to win. Spike enjoys winning. To accomplish this, Spike will play whatever the best build is. Spike will copy strategies off the Internet. Spike will borrow other players’ strategies. To Spike, the thrill of the game is the adrenaline rush of competition. Spike enjoys the stimulation of outplaying the opponent and the glory of victory.

And, finally, Johnny:

Johnny is the creative gamer to whom the game is a form of self-expression. Johnny likes to win, but he wants to win with style. It’s very important to Johnny that he win on his own terms. As such, it's important to Johnny that he's using his own strategies. Playing a game is an opportunity for Johnny to show off his creativity.

Now change the wording to be about coaching basketball, and it's easy to see that George Karl is Johnny:

Johnny is the creative coach to whom basketball is a form of self-expression. Johnny likes to win, but he wants to win with style. It’s very important to Johnny that he win on his own terms. As such, it's important to Johnny that he's using his own lineups, not the traditional PG-SG-SF-PF-C lineup that everyone else uses. Coaching basketball is an opportunity for Johnny to show off his creativity.

If he was playing monopoly, Karl would refuse to buy Park Place or Boardwalk because he wants to win by getting Hotels on Connecticut, Vermont, and Oriental Ave and bleeding you to death slowly. He's that poker player you know who says that he always loses with Aces and that his favorite hand is J-7 of hearts. He's the guy who always takes Gen or Viper in Super Street Fighter IV because he doesn't want to cheese it up with Ryu. If you're playing NBA Live, he's not happy unless he's beating you with the Bobcats.

I'd like to re-iterate that not everything about the Nuggets has been Karl's fault. Lawson has been awful, Gallo has been worse, and the schedule has been brutal. But he simply has not been making the most of it. JaVale McGee should be playing more minutes. Maybe his asthma is a factor but I call bullshit on it being so bad he cannot even play 20 minutes a game. He should be playing 25. Furthermore, if conditioning is an issue, there is no acceptable reason that Koufos doesn't get the minutes that McGee cannot play. And lastly, the nonsense lineups with Miller, Lawson, Iggy, Gallo, and Faried (with Brewer in the mix as well) have just got to stop. Gallo maybe tall but he is not a power forward and never will be. And asking him to play a position he isn't very good at is not the best way to help him overcome his struggles this season. This lineup is particularly bad because it exascerbates Faried's defensive shortcomings. One of the reasons that Faried's fouls are so high is that Karl is often asking him to guard the opposing center, a hopeless assignment.

Which brings me again, back to Gallinari. Why is he being asked to do so much, and being given so much trust? He's had about one season of above average play (the end of 10/11 and the beginning of 11/12 before he was injured). And although he looked very much like a player about to break out into a star, he never completed that break out. His shooting has been all over the place this year, and much of his decision making has seemed suspect (he seems to be passing up lots of open threes, probably from doubts about the aforementioned shooting, and driving into congestion a lot). It would seem logical to try something else, like a lineup with two bigs and Iggy at the the three.

Finally...seriously, why is Brewer being paid to play basketball? For a supposed defensive specialist, he makes an enormous number of defensive bonehead plays per game. In particular, he must lead the league in two categories: committing stupid fouls while trying to make up for a previous mistake, and leaving his defender open for a corner three to help out a teammate that doesn't need the help. And, I'm sorry, but if you are as eye-gorgingly aweful at the offensive side as Brewer is, your defense ought to be perfect for you to get any burn at all. He's one of Denver's worst performers, and that's despite being hotter than the sun from 3-point land for the first ten games. As he continues to regress to the mean, his minutes are going to hurt Denver, a lot. He really shouldn't be playing.

I am convinced the Nuggets have lost a couple of games through this kind of mismanagement, and with the Nuggets playing such a brutal schedule in the beginning of the year, a record like 10-6 or 11-5 would make them one of the best in the West, with lots to look forward to as the schedule gets easier. Instead, they're 8-8 and digging themselves a hole they'll have to climb out of. Future games against weak opponents now have lots of extra pressure because they become "must wins".

But hey, Nuggets fans, when you win, you win with style.

Wp_000390_normal

Luke Gregory Caruana

Based on the new precedent, Karl has got to be expecting a fine for the way he is managing this team. Hope David Stern doesn't see this

199 days ago

Picture?type=square

Brett Gunter

Patrick,

I suspect the unwarranted loyalty Gallinari receives from Mad King George is simply the eyeball test, taken to extremes. When he does something good, he looks absolutely spectacular doing it; I have in mind the insane behind-the-back pass along the baseline that stayed on ESPN highlight reels for a week (if you saw it, you know instantly which one I mean). Karl probably sees flashes of that potential all the time--and ignores the actual lack of production in favor what might be. By the same token, he forgets about the 14 rebounds Faried collects in 22 minutes when he benches him for going over someone's back after a 15th. His continued employment reflects badly on the Nuggets' management.

This isn't just a Wins Produced argument; even if you've never heard of advanced metrics, you can't ignore his flouting of CONVENTIONAL wisdom. Management just shelled out $50 million for a 7-foot C--giving them THREE on the roster, who don't COMBINE for 48 minutes per game! The average minutes per game at C for actual centers goes even lower when you add the fact that, inexplicably, he occasionally plays TWO of them TOGETHER! In contrast, I've seen him play lineups where his SF is his tallest guy, to go against BIG front lines (Gasol/Randolph, et al). I'm going to offend a bunch of people now, but I doubt a coach pulling lineups out of a bag who was not also a cancer survivor would still have this job.

199 days ago

Default_profile_1_normal

Brian Foster

I have to be a nerd for a moment here and nitpick the Monopoly analogy.

Boardwalk and Park Place (dark blue) are more like Jimmy/Timmy properties, they are expensive to development and hard to land on, but if someone does land on there when it is reasonably developed it will crush them. Dark blue is pretty good late game when everyone is developed and you have a lot of money, but situational early and mid game.

Light blue is what Spike would go for in most situations, you can develop them cheaply, and because they get landed on all the time due to the abundance of "advance to go" type of cards they have great ROI, bleeding out the guys going for expensive properties like dark blue so they can't build up a lot of houses. While still being inferior to orange and red in most situations (orange is the Lebron of Monopoly), light blue is a really solid property group.

http://www.amnesta.net/other/monopoly/
http://www.tkcs-collins.com/truman/monopoly/monopoly.shtml

199 days ago

Nba-geek-avatar

Patrick Minton

Omg I must admit I just pulled the monopoly one outta my ass in an attempt to appeal to mainstream readers.

Your deconstruction of my analogy is AWESOME. You should comment more here!

199 days ago

Pbqf23fs9wi23034q8ui_normal

benjamin durham

Tonight was rediculous. to lose because meeksl/jamisonl/howard were amazing is what it is. but when miller/brewer/mosgov get more burn than mcgee or faried thats so much more reprehensable than spurs resting players after a brutal schedule

199 days ago

Picture?type=square

Jack Adams

There was a play in the first half where Corey Brewer doubled MWP in the post with Faried on him, leaving Jodie Meeks wide open for a made 3. Not long after, Brewer made a nice play, poking away the entry pass from the man posting him up. (It went straight to a LAL player, unfortunately). Seemingly inspired by his defensive tenacity and prowess, he instantly forgot everything about guarding his guy, who leaked out to the wing and hit a corner 3. I switched the channel to Shark Tank at the next commercial break.

199 days ago

Picture?type=square

Jack Adams

And Brian Foster is definitely right about the Monopoly thing. There are numerous ways to get sent to jail (Three doubles rolls in a row, several go to jail cards, and the Go to Jail square). The most common roll is 7, of course, with something like 17% chance, landing you most often on community chest. But 6 and 8 land you on orange from the Jail Square, and their added probability is around 28%. The math behind Illinois Avenue requires too long of an explanation for a comment. Also, go for Railroads, especially early in the game. (I like Monopoly).

Do you know where I can get player stats by quarter? I'm currently slogging through Play by Play stuff and logging the stats myself. Thank God for small sample sizes.

199 days ago

Img_3451_normal

Jon C

I totally agree with this article. Karl and other coaches really try to find ways to justify their existence and their multi-million dollar salaries. Playing guys like Faried and McGee minimizes the coaches impact. These players limit the creativity of the offensive play calling and they do most of their best stuff on broken plays anyway. Karl will not get credit for a great play if it results in a missed jumper that Faried rebounds and dunks. McGee does the same thing as he freelances getting by on the potent combination of superior size and athleticism. I think Karl wants to run more interesting plays and continue the illusion that playing different players is a "chess match" where you match your players against the opposing teams players on a lineup by lineup basis while divining who of your group is hot and will play well in the next few minutes. In reality, it might be much more logical to figure out the lineups for optimal rest, set them ahead of the game, and then not deviate (except on rare cases of extreme early foul trouble). Coaching implemented changes during the game are just going to be based on the "eye-test" and the last few minutes of the game (super small sample size). Probably nearly useless changes. But they do get the entire coaching community paid like stars.

198 days ago

Picture?type=square

Jason Lancaster

It's telling that the words "roster balance" don't appear anywhere in this article, as if the Nuggets would be better served to play Faried, McGee, and Koufos simultaneously...a lineup so ludicrous that no coach would even attempt it.

It's all fine and well to talk about the importance of giving minutes to the most productive players, but production doesn't occur in a vacuum. Gallinari's long-range ability opens up spacing for Lawson and McGee, and when Gallinari slides over to the "4", that creates spaces for Faried (or McGee or whomever is playing center) to grab a rebound.

You'll get no argument from me that Brewer is getting too much burn. His defensive lapses ARE a little too numerous - and his shooting much too poor - to justify so much time on the floor. However, it's an 82 game season...I suspect part of the reason Brewer is given playing time is that it forces Jordan Hamilton to work very hard to get minutes.

In closing, Karl is stubborn, but that's true of stat-centric basketball analysis. Roster balance isn't optional.

198 days ago

Picture?type=square

Brett Gunter

Jason,

If you think a Faried/McGee/Koufos frontcourt is absurd, how do you feel about a PF/C pairing of Mozgov and Koufos, apparently a Karl favorite? (I can't say how many minutes they have together--but I've seen it more than once, and the Nuggets are not my top priority to watch.) On one hand, it seems to present many of the same spacing issues; on the other, it implies that Koufos (ny process of elimination) is capable at the 4--and I'm willing to submit to even the eyeball test here and vehemently say Kosta is NO power forward. I'm trying to come up with a less suitable pairing I've ever seen on the court at the 4-5.

(Thinking.)

(Still thinking.)

Old Shaq and older 'Zo? Nah; Mourning had a jump shot.

(Still thinking.)

Sorry, I got nothing--but let's go the other way: Clearly the Madness of King George is such that HE believes Koufos can play PF. It CAN'T be Mozgov (eyeball test again), so he's comfortable with Koufos at PF for some specific length of time. If that's true, why wouldn't Karl put him out there with JaVale at C and Faried at SF? Because none of them can shoot? This is currently true of Gallinari and Lawson, as well; if they are able to space the floor even when unable to hit the broad side of a barn, I would suggest that Faried can miss from 3 just as convincingly.

Don't get me wrong; I'm not arguing those 3 guys need to get extended minutes together. However, I have to take issue with the premise that it's any worse than some of the lineups Karl trots out regularly. As for the idea that Gallo at the 4 helps Faried--on defense, Gallo typically guards the opposition 5 (the conventional wisdom is that Faried's too short, I guess); watching him recently on the low block against Marc Gasol was, um, unique.

If I've missed actual comments from the coach, which explain his reasons behind some of the fantastical groupings he's sent onto the court, I'd be delighted to hear them. I suspect many of the explanations I'm hearing from other people (you may be one such, but I do not assume), claiming to understand his reasons, are the equivalent of blaming a drought on an angry harvest god.

197 days ago

Nba-geek-avatar

Patrick Minton

Jason,

That's quite simply one of the silliest straw man I've ever seen. You know your comment appears right below the actual article, right? Where we can all read through it and see that not once, anywhere, do I hint at playing Faried, Koufos, and McGee together?

Did you not comprehend that my entire article is all about how a traditional, BALANCED, PG-SG-SF-PF-C roster would be better for the Nuggets than the crap Karl keeps throwing at the refrigerator wall, or did you just not bother to read it?

McGee and Koufos combined do not average 48 minutes. How the hell is it "Roster Balance" to not have a center on the floor? How the hell is spending huge stretches of fourth quarters with Gallo, Faried, Dre, Iggy, and Lawson lineups "roster balance"?

Karl's problem is precisely that his roster is almost NEVER balanced. He's playing a small forward out of position at SG, bringing a PG off the bench at SG (and often pairing him with an undersized PG, giving him a super-tiny backcourt), he's playing many lineup variations where Faried is the only one on the floor who can rebound well but asking Faried to defend guys 5 inches taller. And you're saying I don't understand roster balance?

It's hard for me to actually think of a worse way to defend Karl's lineup choices than Roster balance.

197 days ago

Picture?type=square

Adam Tarr

Just for argument's sake, what would a good, balanced minute allocation look like for the Nuggets? I'd say...

McGee and Koufos split 48 minutes at center.
Ty and Miller split 48 minutes at the point.
Iggy and Faried each get ~36 minutes.

That leaves 72 minutes... You could give Anthony Randolph 12 minutes per game backing up Faried (although I do think Gallo can fill in this role if the other team goes small). If Gallo and J-Ham each get around 24, that leaves 12 minutes, which could take the form of more time for Gallo/J-Ham, 2-PG lineups, or some time for the rest of the bench when things are going well.

Seems pretty reasonable to me.

197 days ago

Default_profile_1_normal

Brian Foster

Thanks Patrick, I wish I had more to say, I know next to nothing about math and statistics but I'm crazy enough about sports and games that I'll chip in where I can.

As for your point, Jason, I suggest checking out Ari's recent series of article on Wages of Win about floor stretch and the strategic reasons for playing a balanced lineup. I don't think that the whole series has been published yet, so I'm not sure if the final conclusion will address your specific comments, but at the very least it's related to this conversation:

http://wagesofwins.com/2012/11/06/an-introduction-to-nash-equilibrium-lineups/|

http://wagesofwins.com/2012/11/20/a-beautiful-lineup-nash-equilibrium-lineups-part-ii/

196 days ago

Picture?type=square

Jason Lancaster

@Brett - I don't pretend to understand Karl's rotations, but I don't pretend to know more about the game than a man with a lifetime of experience. It's beyond arrogant for a fanalyst to call out an NBA coach, especially one with Karl's resume.

To be clear, I'm not saying that I think George Karl has all the answers, but I'm quite certain that Patrick Minton - while well intentioned - is absolutely, unequivocally LESS qualified to select the Nuggets rotation. Or to argue that Karl doesn't know how to coach, a ridiculous notion prima facie.

@Patrick - There's a difference between defending Karl's rotations (something I'm not doing) and pointing out that analysis is worthless without considering balance.

In the 4th paragraph, it's said that McGee, Faried, and Koufos are the Nuggets best players. Then, in the following paragraph, it's said that "Clearly, the best players have not been getting the most minutes." If this isn't tantamount to arguing for a line-up of Faried, Koufous, and McGee, I don't know what is.

More to the point, there's a difference between throwing out some lazy analysis on the sidelines and coaching an NBA team for an entire season. I'm no Karl homer, but I'm not an idiot either. I recognize that Karl's line-up decisions aren't *just* about stats and production. They're about rewarding good play in practice, incentivizing certain players to work harder (McGee and Hamilton, for example), and finding the right mix of players that will thrive in the playoffs.

So, in other words, put down the calculator, think about the big picture, and show a little patience. If you think this NBA coaching thing is all about trotting out the best statistical line-up, you're missing the boat.

192 days ago

Picture?type=square

Zachary Young

You can't say that Karl is qualified to select the Nuggets' lineup just because he does select it. That's a circular argument. Patrick has provided actual evidence for his viewpoint, and dismissing his opinion because he's an outsider is an insult to the very idea of debate, since you should be addressing his argument. You can't just dismiss statistical analysis as "missing the boat" and expect to have a meaningful conversation. It reduces us to trading anecdotes that do nothing but reinforce unverifiable preconceptions.

Also, I'm fairly certain that there are 96 minutes of big man basketball available. That means that McGee, Faried, and Koufos could each play 32 minutes without ever having more than two bigs on the court at a time. The whole point of the article is that if you've got multiple good big men (like the Nuggets do), you shouldn't play lineups without big men (like the Nuggets do). Playing "small" lineups is fine if your best players are smaller, but to spurn the traditional PG-SG-SF-PF-C when your roster naturally lines up that way just doesn't make sense.

192 days ago

Default_profile_4_normal

HoopCount

Questionable analysis. Both McGee and Farried score more points more efficiently when the other one is on the bench. There is such a thing as floor spacing, and clogging the court with big men simply because they have great stats, is a recipe for failure. It's important to look at all aspects of the game, not just the numbers, which can only show so much.

191 days ago

Picture?type=square

Zachary Young

Even if there is evidence to back that claim, HoopCount, I'm not sure they've shared the court enough for one to draw any reasonable conclusions. I'd also say that since Wins Produced can show 95% of what goes into a win, any negative effect a lineup has will actually show up in the numbers after its been tried over a large enough timeframe. Playing guys that we know stink because the good players might have trouble together is silly.

Here's a question: where's the evidence that "clogging the court" isn't the basketball equivalent of baseball's "clogging the bases," as in a really good thing to do if you can? I'd think the burden of proof would be on the people arguing against putting better players on the floor. If I've been reading the Nash tables at Wages of Wins correctly, so long as somebody is stretching the floor a little bit - I'd hope an NBA coach could manage that with three perimeter players - and somebody is rebounding - any lineup with Faried and McGee will do that just fine - you should just play your best players as much as possible.

190 days ago

Default_profile_4_normal

HoopCount

There is evidence (via NBA.com), and they have about 130 minutes of playing together, compared to 330 apart.

The issue isn't wins produced, but with the mindset. From the chart and article, it seems the author was advocating placing the 5 best players ( in terms of WP) together would undoubtedly produce the best possible results. That's just ridiculous. If I had 5 Dwight Howard's on my team, they would all have great WP numbers, but would be almost useless if played all at once. Things like floor spacing and diminishing returns are important, and just because they can't be reduced to a number, doesn't mean they aren't real.

190 days ago

Picture?type=square

Zachary Young

I hope that you can see that 130 minutes isn't a tremendous sample, especially when the figures will also be affected by the other teammates, as well as the opposition.

That's not the argument at all, so don't go throwing up straw men. Patrick specifically argued in favor of the traditional PG-SG-SF-PF-C lineup in the article, and again here in the comments. The best players should get as many minutes as possible (a position which shouldn't need defending), Karl's lineups don't do that, and he can't even claim that they are necessary from a lineup balance perspective because he's not giving enough minutes to front court players. Read the paragraph that begins "But what is surprising is that no center..." and you'll get an answer to your concern.

Ari's recent floor stretch articles on Wages of Wins are also really pertinent to this conversation, and demonstrate both that the Wins Produced community understands basketball and that math can, too.

190 days ago

Default_profile_4_normal

HoopCount

His argument wasn't promoting traditional lineups, it was promoting the play of the 5 most efficient players at once. He used the lineup issue to illustrate a point.

There is a difference between the most productive (Wins Produced), and best for the team. You won't catch me supporting George Karl's roster moves, but what is being proposed here is equally suspect. The 5 most productive players on a team are usually that productive because they have the necessary system around them. Big men play well because they have the space to operate. If you take all 3 of your big men and put them on the floor at the same time, they won't be nearly as productive. Their rebounds will go down ( stolen from each other). The paint will be packed with defenders, decreasing both scoring attempts, and scoring efficiency. Then we'll have another article on how all 3 of these big men need to be benched because they have low WP's, and the cycle will repeat itself.



190 days ago

Default_profile_1_normal

Shawn Furyan

Wow, I'm not sure I've ever seen someone double down on an argument that had already been called out as a straw man. Congrats on your testicular immensity HoopCount. Having had several conversations with Patrick, I'm fairly confident that he's not quite so dumb as to espouse the argument that you are suggesting that he is, especially when he's already said that he's not espousing such an argument. Further, this is an argument that has been explored extensively and might be boiled down to be called the 5 Dennis Rodmans argument. Nobody thinks that throwing players together willy-nilly without regard to position or skill interactions is a winning strategy, and while it's entertaining to watch you assert that there is a person who holds such a ridiculous belief, doing so is just not a productive exercise.

189 days ago

Default_profile_4_normal

HoopCount

The point of this article, was that the Nuggets weren't playing their best 5 (in terms of WP). The author said the Nuggets need to play their best 5 as much as possible (again, as determined by WP). This is tantamount to saying the Nuggets should play Faried, McGee, and Koufos together for at least a large part of the game. As I illustrated in the previous posts, this is a ridiculous lineup that would likely make things worse for the Nuggets. He did everything he could to say it without saying it, and whether he intended to or not, fooled no one.

Other posters on this article expressed similar sentiments, saying multiple one-dimensional big men on the court would do just fine (because they have goof WP's), and that the idea of balancing the floor is comical, because "those" players were bad (in terms of WP). I don't have a problem with WP, or the premise of the article, but simply throwing out those numbers to determine who starts is wrong.

189 days ago

Picture?type=square

Zachary Young

Except it is NOT tantamount to saying Faried, McGee, and Koufos need to play together because the length of the game and normal rotation minutes mean that a competent coach could get them all 30+ minutes without ever having more than two of them on the floor at a time. Starter-type minutes is a misnomer because the sheer number of lineup changes in basketball means that a high minute total doesn't actually necessitate starting. Let me illustrate this as clearly and as simply as a I can:

Faried plays 32 minutes at PF. He is joined by Koufos for the first 16 minutes and McGee for the second 16 minutes. When Faried leaves the game after reaching 32 minutes, Koufos replaces him and joins McGee for the final 16 minutes. Everyone gets their minutes and no lineup engages in any kind of strangeness.

While obviously the actual units of time would be divided up in a more complex manner than in my example, it should be clear that since there are 96 minutes of big man basketball available in the traditional lineup, only giving about 70 to your three best big men (and then mostly giving the remaining minutes to wings) doesn't make sense.

189 days ago

Nba-geek-avatar

Patrick Minton

Jason,

If you are going to appeal to authority here then we're essentially done. The entire premise of my blog is that the major decision makers in the NBA are almost all doing it wrong.

Years of experience doing something poorly does not make you an expert.

188 days ago

Nba-geek-avatar

Patrick Minton

HoopCount,

I have no words. I'll repeat myself that I do not believe that McGee, Faried and Koufus should share much floor time together. I'll repeat myself yet again and claim that even if Karl is forced to play "Get the man, chicken, and wolf across the river" with those three players, he could still get his best 5 players more playing time.

"Play your best 5" and "Play your best 5 all at the same time no matter what position they play" are two sentences with different meanings. You tacked on the clause in italics, I did not.

188 days ago

Picture?type=square

Jason Lancaster

@Patrick Talk about circular logic! I say Karl is at least a *decent* coach because:

- he's recognized as such by nearly everyone in the industry...I'm sure you're say that "everyone is wrong but me," but that smacks of arrogance more than reason
- his teams have performed OK against similar talent in the playoffs
- he's managed to get some great performances out of under-achieving teams in the past

Your argument is (essentially): My math is more qualified to coach an NBA team than a seasoned professional, and the proof is that my math always shows Karl can't select rotations.

For the record, I don't think Karl is much above average. Time and again we've seen puzzling rotations, a stubborn insistence upon sticking with players who "play the right way" rather than playing the most talented people, and -- most maddeningly -- Karl has botched relationships with dozens of players. I'd argue relationship building is the coach's most important job, and he's clearly not that great at it.

Still, sitting here with a spreadsheet and calling out rotations without acknowledging all the other factors (rewarding players who practice hard, giving everyone a chance to prove they can or can't contribute, trying things even though they probably won't work) is ludicrous.

I like what you produce here, but it's the foolish to assume your mathematical model is beyond reproach. It's been shown that WP can't predict outcomes any better than any other advanced stat, and none of them are very good. (I'd post the links to this data but I'm sure you've seen it.)

Thanks for responding.

188 days ago


You must be signed in to leave a comment.