I know that right now, actual basketball being played on an actual court with actual NBA players is probably pretty far back in Thunder fans’ minds. I think the Thunder fan thought flow right now goes: “Draft-Draft-James Harden-Ricky Rubio-Food–Work–Draft-Sleep-Trades-Food-Draft-NBA Finals-Hasheem Thabeet.” Or something like that.
But the truth is, the Finals are what we all want. I may be totally crazy here, but I’d much rather have won a title than won the lottery. Heck, I’d rather have gotten swept in the Western quarterfinals than been in the lottery. But that’s me. I’d rather be playing than watching and wishing. The Finals are the reason we all care so much about free agency and the draft. Because you know, we want players that can help us get there.
Anyway, so with the Finals starting this week and I was thinking: What can Thunder fans and/or players take out of this? Here’s five things:
1. A superstar is needed, but the other four guys are just as important. Watch the way the Lakers maneuver with Kobe Bryant. He’s the unquestioned MAN of that team, but there’s a reason they are here and the Cavs are not. Those reasons are Andrew Bynum, Trevor Ariza, Lamar Odom and Pau Gasol. Kevin Durant is already a fantastic - I say FANTASTIC - player, but he can’t do it on his own. Nor could Russell Westbrook or Jeff Green. It’s got to be five guys working as one. Or at least three or four. Kobe doesn’t have to take every shot. He’s perfectly content scoring 18 on 6-11 shooting. He just waits and strikes when needed. Nobody does it better.
I remember some of my favorite basketball the Thunder played this season was while KD was out (don’t read anything into that). The team swung the ball side-to-side, made crisp cuts and moved the ball everywhere. It was beautiful basketball. They didn’t have the out of KD standing on the wing. They couldn’t just toss it to him and hope for the best. They all had to contribute. That’s the type of basketball is really takes to win, but insert a superstar player into the middle of it and you’ve got something special. It’s harder to figure out than you think though. The Magic understand Dwight Howard’s abilities and deficiencies and adjust accordingly. Same with the Lakers. You have to understand how everything works around your best player. Both these squads have that figured out pretty well.
2. You don’t need a great point guard – you need players that work together. Look at both the Lakers and Magic – do Rafer Altson or Derek Fisher really scream, “GREAT POINT GUARD!” to you? They both know their role and handle it well. They hit shots, pass well, can score a little but most of all manage the game. I just wrote about Russell Westbrook’s issues about managing the game. He could learn a lot from either of these point guards and neither one is really all that great. But they don’t try and do too much. They just play within the flow, take shots when they have them and try and make their teammates better. That doesn’t mean you’re supposed to just take four shots and average seven points per game. OKC needs Westbrook’s scoring and playmaking ability. Heck, Alston had 26 in Game 5 against Cleveland. But he chose his spots. He wasn’t forcing anything. He was just playing. I really think Westbrook will start to figure that out this season. I really do. And that makes my face tingle.
3. Defense really does make the difference. The Magic aren’t known for defense, but boy, they guard you. Every possession was an absolutely grind for Cleveland. Orlando is physical, in your face and don’t just give you an easy two. Ever. They truly make you earn it. The Lakers are more your classic defensive team. They guard well, help defend well and get in passing lanes. They don’t do anything outrageously awesome, but they do everything well. Lots of teams can score the ball. But the teams that make it more difficult than usual are the ones that go far. Stan Van Gundy convinced the Orlando of it this postseason. And you can see where it’s got them.
4. You need an identity. What was the Thunder’s calling card? And don’t say “losing” because that’s not what I mean. When you see San Antonio, you know what they do. Same with the Suns or the Nuggets. Good teams have an identity. Great teams have one and do whatever it is extremely well. The Lakers play solid defense, run the triangle offense and have Kobe. The Magic play excellent defense, shoot the three and pound it down low to Howard. You know what they are going to do before the game starts. But they do it so well that even though you know it’s coming, you can’t stop it. What’s going to be OKC’s identity? Defense? Run and gun offense? Solid defense with efficient offense? I don’t think even Scott Brooks really knows yet, though I know that they want a good defensive team. But then again, who doesn’t? Oh yeah, Don Nelson.
5. When all else fails and goes to crap, you need someone that can put you on their shoulders. And in the case of Dwight Howard, some extremely wide and broad shoulders. But the Lakers have Kobe. He won me over when he walked off the bench in OKC after the Thunder had pulled to within a few points and immediately ripped off 12 straight. It was the ultimate, “I got this” moment. And he can do that. He takes the biggest shots and makes them. He’s ready and willing to take the game over and more importantly, he’s not afraid to fail trying. Dwight Howard had his moment Game 6 against Boston after calling out his coach. He had it for his team. Kevin Durant will absolutely get there. He had some moments this year, but sometimes would disappear late in games. But for the 476th time, he’s just 20 and he’ll figure it out and sense the moment. And when he does, it will make us very happy.
Oh, and just because I know you care – Lakers in 6. What say you?
If you compared the sums of the 5 players adjusted to the adjusted for a lineup for all lineups and found the correlation you could say more about the persistence of the player adjusted and you might be able to find out more about player impact on other players and total team results for types of players. Is you were really serious about the game.
Should be
Follow the lead of the very best you’d increasing that to 3-4 times "if you get to the playoffs". "And want to practice playing that way in preparation."
And
Most coaches aren’t "modest enough about reading context with precision a lot of the time". "That the very best teams with probably among the very best coaches just roll with a few lineups for a large share of the time in the playoffs does show more modesty though. Appropriately so."
Thunder used 405 lineups in mostly small to very small amounts. Was that the best strategy? I don't think so. Apart from what the stats and the logical way to get the most out of the team based on what they say, playing guys with other guys to find the best performing combinations that you intend to use in the future when it matters is good experience. You got to learn what teammates will do when with each other in a certain lineup.
If you question my ability to judge and tout or criticize the very most used / tested lineups based on the stats are you at the same time comfortable that Scott Brooks knows which of 405 lineups (almost all played lightly to very lightly) to use in any given situation? I'll take a strategic stat-based assessment of the overall performance of top lineups (or even better a split based on similarity of the opponents previously faced to the current one with the stats of most similar teams weighted favorer than those of those less similar) over most NBA coaches situational tactical decision-making. I am being modest about my ability to read the right lineup. Most coaches aren't.
Importance of 5 man lineups and the top lineups being used a lot and being good? Check this out by the 4 conference finalists in the playoffs:
Denver just top 2 lineups 40% of time
Lakers top 2 lineups 45% of time
Cavs top 2 lineups 47% of time
Magic top 2 lineups 36% of time.
Different context but Thunder in regular season top 2 lineups were used 11% of time. Follow the lead of the very best you'd increasing that to 3-4 times. Those lineups and the next 5 better be good if you want to win in the playoffs.
Boston last season was a couple of all-around perimeter players, a PF who was high on bringing non-scoring value (and some scoring too of course) and a defensive center. Orlando and LA this season are more like that than different. It is not that surprising but if you study the details you will see a lot of teams doing something significantly different here and / or there. And generally less well. The best GMs old-school or new tend to end up doing the same things.
Just quickly checking Durant is kinda high on being a pure scorer and overall this type is not very valuable. A scoring perimeter type who does more than just score is on average more valuable. Durant is half and half right now.
Kristic is the least valuable -from an overall team results standpoint- on average type of big- an interior scorer (who really does it with jump shots mostly).
Collison is pretty high on a very valuable type (with an awkward name of scorer's opposite).
Green is a very hard to classify mixture that isn't strong one thing or another. Some "mixed" players are great, others aren't much.
The newer guys I could guess their types but haven't seen the numbers run.
If interested in types check out this thread
http://sonicscentral.com/apbrmetrics/viewtopic.php?t=1833&highlight=type+lineup
It includes my early efforts to identify the best type sequence for a lineup.Boston in 2008 had it right.
And the sample size argument is actually a good reason for doing what I recommend. Play your best guess lineups more and you'll get more meaningful data. More meaningful data can leads to better choices and better results. Continue to scatter team minutes into hundreds and hundreds of small minute lineups and you never really learn anything solid and are just relying on coaching experience / instinct. Coaching experience / instinct is rarely analyzed comprehensively in depth quantitatively. They can't escape game to game review or ultimate season record review but they are often escaping review for playing a bad lineup too much or not playing possibly good ones enough. That is why I look at raw and adjusted lineup data and player pairs, etc.
A strong performing lineup is a real hammer. A strong player in some part of the game can be too but is easier to still beat unless the rest is taken care of... and produces a strong lineup.
I did study 5 man lineups using a typology developed by David Sparks awhile ago and some sequences of types did much better. But to be more meaningful you'd need to know type, quality, offensive/defensive split, consistency / leadership, etc. There is plenty to mine, if there was too time and reason.
For the record I discussed 9 teams, half of them in depth and cited several types of league level data. Against my assertion that 5 man performance matters and the best available performance data (lineup adjusted) should be considered (more) in optimizing the lineup rotation the arguments boil down to the data isn't that good because of sample size and noise or having a rotation is enough. I conceded it is not clear-cut, one sided but I know where I stand. I guess others do too, so we're probably done.
should be sometimes my writeups might "not" be real strong on everything
but I would say I don't say or mean everything that some say I say or think I say.
You probably won't be altered by what I wrote Vince. I'll be glad to admit that lineup adjusted data is not rock solid or unarguable. But I will say it is better to consider than not to. I could write more but I have other things calling.
But Keith deserves a reply too.
"If you don’t mind my asking for more information, how does that work against individual performance.'
I guess I would say that players have both direct and indirect offensive and defensive impacts. Against counterpoarts and as part of a team. The best way to measure all these factors would be to run adjusted +/- down to the 4 factor level (direct and indirect impact on FG%, turnovers, offensive rebounding and getting to the line on offense and defense). 16 factors. It could be done I have advocated it but I am not the computer whiz to do it and haven't convinced the few who can to do it in public for free. I don't know if any team does it but if I ran a team I would.
"Several people charted in several games how Westbrook actually played rather poor defense, and it was obvious how inefficient of a scorer and distributor he was. How then did he end up as an overall positive when Green and Durant - players who could do at least one thing well - end up negatives?"
Don't have the adjusted 4 factor data I just described but making guesses from what is available I'd say that Westbrook's penetration and passing had some impact of freeing others on offense but Westbrook's biggest impact was offensive rebounding. On defense he was ok but not great. Early on I think folks didn't know what he could do and how he played. As they learned and he got tired and maybe he focused to much on individual offensive numbers his defense faded.
For Durant his individual defense wasn't terrible but team defense was with him. His help defense is probably a good part of that but it is not all on him. Green got better but was still weak on team defense compared to where you want to be.
"Also, do you feel that it is a personal issue when saying certain players don’t work well together, or a type issue.
I think it is both. I have worked a little with types and would do more if I was being paid for it and a team was actually using it. I might do something with typoes later next season after the new guys show what they can do more.
on Green and Durant... "do they play differently when together/apart?"
They don't change their individual offensive games when together hardly at all but the team defense really dives with both compared to one on the court.
A fellow named Ryan Parker studied player impact on team shot charts but it was for last season. If he updates it it might be quite helpful for understanding a fuller list of player offensive and defensive impacts. Green and Durant last season helped increase our and opponent shots in the paint but with opponent increases were twice as big.
"For a 5 man lineup, is there something we can take from the players who worked well together, or is it something more mental with the players.
For example, Mason and Thabo seem to have very similar games to me. They focus on defense and are poor range shooters. Does swapping them make a considerable difference on the lineup?"
I was surprised how different the results were with Mason vs Thabo. Type is only part of it.
It is real speculative here. I would guess that Mason and Watson did well with the other best 4 in part because they were veterans. Maybe Thabo plays more like more vet like. more like Mason when he has played with these guys longer and they with him but there wasn't great chemistry between Thabo and his teammates in Chicago. He is all about team but that doesn't mean he actually plays better with others. He doesn't on offense. Defense it is so far a mixed bag. May be too aggressive at times.
Thanks for the challenging dialogue. I don't have all the answers and sometimes my writeups might be real strong on everything. I am too used to not getting responses with some of this stuff. But I can't write up everything without turning many off.
"Again, given sample size, one bad quarter (or one or two good opponents) dramatically changes the results."
One bad quarter could affect a sample size below 100 minutes a significant amount. That is why I push for teams to use their best lineups more to get meaningful results.
But one or two good opponents would not be an issue unaddressed as these results at basketballvalue.com are adjusted for teammate and opponent quality (albeit imperfectly)
"You’re only proving that they are correlated, not that Jones’ addition necessarily causes the decline."
I should amend the impression I gave- I am not necessarily blaming Dahntay Jones for it all, though a simple reading of lineup adjusted vs the sum of player adjusted points the finger at him heavily. I am saying that putting him in the 5 man lineup dynamic appears to have huge consequences compared to the other 4 plus JR Smith and that maybe a coaching mistake to spend as much time with him in that lineup if you have better alternatives. Smith could have played at least 6-8 more minutes and little differences can add up.
"The most used lineups cannot, on a per-minute basis, be better than the least used, relative to team performance."
They are. I tried to re-find the link but haven't yet. You go to concede the point after rejecting it. It isn't surprising that it is true. It is surprising that given that is true that coaches fritter away 2/3 - 75% of their time on other less proven, less productive lineups.
A 5 man lineup is really a set of 5 players directly and 10 player interaction pairs. Changing between Jones and Smith alternative 1 of the 5 players and 4 of 10 player pairs. You could argue that changes at least 30% of the total of what affect lineup production in this team game.
Some other examples I’ve written about elsewhere:
Boston hobbled by PF injury went with Davis in with the starters big time- most used lineup in the playoffs by any team by over 40% and most used by the Celtics by a ratio of 5.5 to 1. Trouble was though is it was -7 per 48 in the regular season and stayed that way at -8 in the playoffs. Based on what I see Boston might have been wise to try to go small more often.
Rick Adelman certainly got on the lineup concentration wagon, after the challenges and choices of the bulk of the season were it was not highly concentrated, using his most used lineup about 40% of the first round series' total minutes. Nate McMillan? Most used lineup was about 13% of total minutes (probably on low end for playoff teams) and in line with regular season. It didn't work against Houston.
Blake-Fernandez-Roy-Aldridge-Pryzbilla was the 6th most used lineup in regular season - and at +33 per 48 minutes- but at a ridiculous 1.5 minutes a game average. Nate turns to it in playoffs, but too late. Didn't know it, trust it enough? Didn't try it enough when he could have.
Roy and Fernandez are each other's best raw player pairs. Roy's net counterpart point and PER advantage at 82 games is twice as big at SF as at SG and using their position assignment data it was a not too small 1000 minute sample at SF. Seems like a pretty strong surface case for using him more there. Houston was beefy at SF but not especially tall or inside oriented. Would Roy at SF a lot more with Rudy have worked? I think it was more likely than not to work better than the starting lineup with Batum. One good half by Artest in game 6 doesn't necessarily mean this was a bad idea. Overall by one crude tool he had a winning matchup record against Artest this season. They needed something different faster.
What other evidence do we have? The Blazers had 11 lineups that weren't negative in the playoffs and Fernandez was in 7 of them (I think most on team). Roy-Fernandez represented about 55% of all the non-negative minutes yet was in only a bit over 1/3rd the total minutes. So it "worked" about 50% more often than other stuff worked. Run Roy- Fernandez 5-10 minutes more a game and the full starting lineup with them suggested above more and maybe the lineups with Oden and Outlaw with them that worked in really short minutes more too and would it have made the difference? I think it could have a much better shot. We'll probably see more Roy-Fernandez next year, if not the best 5 man lineups with them. Hard for me to see why it wasn't done this time, quicker / bigger though. Blake-Fernandez-Roy-Outlaw-Oden "wins" as most positive regardless of minutes lineup +8 in 9 minutes of play in post-season but unless I missed it in the fine print at basketballvalue it didn't see a single second of play in the regular season. Nada. They used that quad with Sergio a bit and it was +18 per 48. A clue worth further investigation if you are paying attention, trying to mine ideas and will to investigate further. There are plenty of opportunities to learn and benefit if you have a strategic plan for minutes use.
Fernandez-Roy-Outlaw-Oden lineups represented 15% of the instances of a positive lineup in the playoffs. It was used about 8% of total minutes, so it outperformed at a rate of 2 to 1 compared to the rest of combos.
Regular season Fernandez-Roy-Outlaw-Oden was almost +20 raw +/- per 48. Just used 78 minutes so you can't rely on that. But why only 78 minutes in 82 games? Focus on your best guess 30-50 lineups and cut out a lot of the tail end stuff that eats up too much of most coaches minute usage in the regular season playing match-up or simply not really valuing lineups are unique production units.
Ah that Blake-Fernandez-Roy-Aldridge-Pryzbilla lineup that rated real good? Turns out it is the best lineup in the league used over 100 minutes.
All the thousands of lineups the league's coaches use? Only 72 were used 100+ minutes and were +5 or better, raw. The top 8 playoff teams or 27% of teams have 38 of the 72 best performing lineups used over 100 minutes. Lakers have 8 of these, Cavs, Magic and Nuggets 5. Those are your 4 conference finalists. Chicken or egg? they go together.
With good performing lineups so rare you'd think coaches would try the better ones more? At least I'd think so. Maybe it is hard to make them work longer in many cases but I'd sure try. I'd think that would have a better chance than what they are doing, using what isn't working a lot (too much) and using many many lineups in microscopic amounts.
Of these 72 only 33 got used over 2 minutes a game, just 19 over 3 minutes a game, and it is down to a mere 14 getting more than 4 minutes a game. Highly suspect lineup management to me. Take the best lineups of a certain minute size after 1/3rd to 1/2 the season and increase many of them several fold depending of how much you like the lineup and trust the data. And certainly use all the results for planning next season.
Also I don't think it is unreasonable to say top lineups probably help decide a lot of the close games at the end.
@Vince
1) Correlation is part of understanding. Scientists do not know the specific process of cancer creation in human cells for 99% of cases. Because of this, they can't say that certain things definitely cause cancer. However, from lots of correlating data, they can give you a pretty darn good idea of what can cause it. By the way, don't smoke. To say that you can't use data because it only shows correlation is to disregard most of the data that can be gathered.
2) The point made is that good teams tend to play their best performing lineups much more often than other lineups. Saying that all most-used lineups use a teams best players is to miss the point. Your 5 most talented players are not necessarily the best 5 people to have on the floor together. For example, Allen Iverson and anyone. It's the reason an over the hill Jason Kidd who can't shoot well is still sought after. He's not the best individual player at all anymore, but his presence and talents make everyone better around him.
3) Data is essential to understanding. Our simple observation as fans is still data gathering. Crow's data is important because it helps quantify those things we think or estimate from watching. It also helps us reconsider misconceptions. As fans, we see certain things and say, "Wow, that is a great play, that guy is really good." Unfortunately, many of us can miss how that great play was most luck, or how that great player is more flash than substance. The whole point of the data collected at 82games, and here relayed by Crow, is to help us see those instances more clearly.
In order:
1. Showing that one lineup from one team involving one player doesn't prove your point. Correlation does not equal causality. You're only proving that they are correlated, not that Jones' addition necessarily causes the decline. Again, given sample size, one bad quarter (or one or two good opponents) dramatically changes the results. So you can't take a season-long sample and use that information effectively in a 7-game playoff.
2. The most used lineups cannot, on a per-minute basis, be better than the least used, relative to team performance. It's close to a zero-sum game, performance wise. If you're saying that teams do better with their most-used lineups, that might have something to do with the fact that their most-used lineups consist of their best players. I'm not sure why we need 82games.com to figure that out.
3. Saying the data is important and "absolutely essential" is meaningless without some sort of backup.
Statisticians like to point out that the murder rate is closely correlated to the consumption of ice cream. Does ice cream cause murder? Of course not. Correlation does not equal casuality. You're proving a slight correlation with small sample sizes but you've done show any meaningful cause and effect or lesson to take for the future.
Crow :
But it is not correct that “all the various combinations of those 8 are going to look good”. I showed that with Denver’s four best and a clunker, not as good on defense as claimed Dahntay Jones. And that is one of many possible examples.
And as for consistency and better performing lineups not correlated, a comprehensive study of all teams and all lineups in the league at APBRmetrics showed that the most used lineups were indeed demonstrably better than the lightly used on a pretty clearcut curve from high-use being better to the lightly used tail of lineups being far worse on average.
5 man lineup and player pair data is very important, to me. If you don’t buy it that is fine. I think it is absolutely essential.
Wow Crow, great stuff. If you don't mind my asking for more information, how does that work against individual performance. Several people charted in several games how Westbrook actually played rather poor defense, and it was obvious how inefficient of a scorer and distributor he was. How then did he end up as an overall positive when Green and Durant - players who could do at least one thing well - end up negatives?
Also, do you feel that it is a personal issue when saying certain players don't work well together, or a type issue. For example, are Green and Durant a bad pair because of their non-complementary strengths/weaknesses as players, or do they play differently when together/apart? For a 5 man lineup, is there something we can take from the players who worked well together, or is it something more mental with the players. For example, Mason and Thabo seem to have very similar games to me. They focus on defense and are poor range shooters. Does swapping them make a considerable difference on the lineup?
A good 8=9 rotation is still important in the playoffs of course,
but top lineup court time increases significantly in the playoffs and the final outcome is increasingly about your top lineups against theirs. If you don't have top performing lineups that do well against other top lineups you are not going to go far when it really matters.
Atlanta and Utah are two more examples of playoff teams who have the wrong most used top lineup, being -4 for the entire regular season on adjusted. Even if we can't be sure they were "truly", fundamentally, negative the odds are better than 2 to 1 they were and the odds they are actually good- say +2 or 4 or better- and just unluckly was 20% or 10% or less.
Of the 6 playoff teams without a top 50 most used in the league regular season lineup that was +5 or better per 48 minutes, 4 lost in the first round and the 2 that won beat teams also without such a lineup.
6 of the top 7 top most used lineups that were +9 or better per 48 minutes made it to the second round or better with one having to be eliminated as being the only such team (Portland) to face a first round opponent with the same attribute.
Lineups play basketball. Rotations and teams don't, directly. Only thru 5 man lineups. That work or don't.
Anon you are right that on a very good or great team combinations of the top 8 are likely to still do well. 3 or 4 of the best players and filler can work. (I should have noted that and was going to revise it even before I saw your note, though it is far less likely on a weak team including the Thunder, the original focus.)
But it is not correct that "all the various combinations of those 8 are going to look good". I showed that with Denver's four best and a clunker, not as good on defense as claimed Dahntay Jones. And that is one of many possible examples.
And as for consistency and better performing lineups not correlated, a comprehensive study of all teams and all lineups in the league at APBRmetrics showed that the most used lineups were indeed demonstrably better than the lightly used on a pretty clearcut curve from high-use being better to the lightly used tail of lineups being far worse on average.
Big picture, coaches show some signs of seeing the best but still way overuse the weak lineups. The Thunder's top 7 lineups accounted for barely more than 25% of total minutes. Almost 75% of the entire season went to lineups used under 80 minutes for the entire season. To make a 100 minute test of a lineup only takes 20 games for 5 minutes each. Only 6 lineups could even pass that very easy to arrange standard even with injuries, acquistions, etc. And four of the six choices to try that much sucked.
"The issue is not finding solid 5 man lineups, it’s finding a solid 8-9 man rotation." I disagree strongly. Finding a solid 8-9 man rotation is merely a starting point. It is far far from the full job complete. That thinking it what a lot of the league shows- to their detriment. Players are not just plug-ins working the same with everyone. Some players are much better or worse with certain other players. Beyond noise in my estimate though it is hard to tell for certain when coaches use lineups generally very diffusely and lightly- a hard to justify pattern To get better data you can rely on more, use your best lineups more. 5 man lineup and player pair data is very important, to me. If you don't buy it that is fine. I think it is absolutely essential.
Anonymous :The issue is not finding solid 5 man lineups, it’s finding a solid 8-9 man rotation.
I agree with that.
Crow is definitely showing next level thinking, but we're obviously building and trying to find what will work best. We ultimately need to find the best 5 man lineups, but right now we're working on: RW decision making skills/jump shot, Livingston's knee, Thabo's jumpshot, DJ White's growth on his face, Desmond's injury, coaching change, we were finding out who can't fit for this team (Wilkins/Atkins/Watson/Swift/Wilcox/Petro).
Brooks had a lot on his plate this year, and I think it was more about development then it was about really getting the 5 man lineups down. Crow's right in this being one of the next steps though. I wish we were there already.
And notice I agreed with your response when I said
"Got to find 5 man lineups that work and then use them a lot."
Not the other way around as you hinted I had or misunderstood what I actually said.
should be "nearly 4000 team minutes" in above
What is silly is playing nearly 400) team minutes and only playing 6 lineups more than 100 minutes (or 1.25 minutes per game) and none above 300 minutes (less than 4 minutes a game). If you want to criticize small sample I'll agree to that and say that then for the entire season next to nothing was learned about 5 man lineup performance that can be used to build on for the future. And then I'll ask why the coach and the GM spent the season that way.
You are welcome to consider it silly Vince. I am simply checking to see what it says and considering it.
I didn't say LA/ORL were successful this year because they used more consistent lineups. As you say it’s more likely the reverse. Those lineups did work, so they used them a lot.
But by doing so they helped their cause compared to teams who didn't use their best lineups as much.
Denver's most used lineup with Dahntay Jones was actually a clunker at -1.6 regular season and -3 in the playoffs and that contributed to their defeat. Denver's top 3 lineups were used 45% less in the regular season than the Lakers. So they were less tested or proven. Of the other two one didnt work against the Lakers and the other was hardly used. Not what you'd expect or want in the conference finals.
Top lineups and top lineup performance matters and few have studies them enough. Even among the confident insiders. In my opinion having looked at lineup results a fair amount.
@Crow
That lineup stuff is kinda silly, quite honestly. There were only 6 lineups that played together over 100 minutes. You can't make any conclusions from sample sizes that small.
And drawing conclusions from the Lakers/Magic numbers -- of course, their lineups do better, because their teams are better. And they used those lineups because they have set rotations and star players. You can't prove that LA/ORL were successful this year because they used more consistent lineups == it's more likely the reverse. Those lineups did work, so they used them a lot.
If Thabo and Krstic are likely subs when the design is complete or temps gone when it is by trade or just not resigning then criticism of this year's lineups was premature. Though hard to avoid doing.
But on the bright side you could say that despite Durant-Green still being overall a bad player pair you can make it work with Collison and Westbrook and the right somebody. But it is pulling a slot machine and you need to get it all right to get the payoff. 3 or 4 right isn't enough.
Best three lineups over 30 minutes- all had Weaver. Durant, Green, Kristic and Collison in 2, Westbrook and Thabo just 1. SG is definitely a riddle to be solved. PG maybe too.
On the topic of "what we can learn" I'll add that the best teams not only have 5+ guys they have 5 man lineups that work well.
Lakers top 7 lineups are all adjusted +/- positive. Magic 6 0f 7. Thunder 2 of 7.
And the Lakers top 3 lineups are used reliably 125% more than the Thunder's and the Magic's top 3 are used 65% more. Got to find 5 man lineups that work and then use them a lot. The 2 Thunder lineups that worked in the top 7 used were Collison - Durant- Green- Mason- Westbrook and Collison - Durant - Green - Watson - Westbrook. Odds are near 100% neither will be around in 2010-11 and only 1 likely for at least part of 2009-10 and then only if a Watson trade doesn't happen or til it does.
You might notice both good lineups were Collison - Durant- Green- Westbrook.
Add Thabo and sub Krstic and you go from positive to -18 on adjusted. Presti full by hi s hand starting lineup at this moment was absolutely terrible early on. Wait n see how it does next year but it is main reason I question the decisions to date. Of course the current starting lineup i snot guarantee to be that next year or beyond.
Add just Thabo and it is still -18 on adjusted.
Take the 4 who did well with Mason or Watson and substitute in Weaver and unfortunately that was terrible too.
Substitute in Livingston and they were neutral but in only 11 minutes so can't really even guess.
Presti has to find the right last piece or start over to some degree. Keeping Mason or Watson is unlikely the answer but they prove you can either solve the puzzle- or occasionally get lucky.
I'd rather not get into the whole Seattle/OKC thing. It's almost a year in the past and we've all done our best to move on. The situation was awful for both sides, but it is what it is now.
I apologize for not taking care of this earlier. My internet has been a mess today and just now have had a chance to look. If any of you have an issue with my deleting comments or you want to continue the discussion off the site, feel free to email me at royceyoung41@gmail.com.
Chad Ford says that the Thunder might be interested in Ben Gordon. I guarantee that the Thunder won't sign him. This is just an example of the all-knowing Presti covering all of his bases. I can hear the conversation with Ben Gordon's agent right now...
Presti: "Okay, so what are you looking for as far as money goes?"
Ben's agent: "We want a five year, 50$ million dollar deal, and we want at least sixty shots a game, every game."
Presti: "Uh... No thanks."
*hangs up.*
@Ty
Apparently, Presti likes B.J. Mullens with the fifteenth pick.
Off topic like always.... We are reported to be interested in trading our 25th pick with cash to Detroit for their 15th pick and a point guard who i don't know the name. Maybe we could pick up Earl Clark or Austin Daye. Austin is a scoring threat, but is weak sh*t. If we can bulk him up maybe he will be good down the road???
The Thunder's collapse has already happened. They are getting better and better, and they will soon be a major contender for years to come. Also, getting rid of Allen and Lewis was actually a good move. With them, the Sonics were a completely average team that would be a border-line playoff team, but never a championship team. They may have had two really bad years after dumping them, but it is all for the best because they got a lot of good players with those high draft picks, and those players will be leading them back to the playoffs very soon, possibly even next season.
@Steve H
Haha. Amen. Presti has the media's head spinning with his interests. People can't keep up. He's awesome.
I think Rashard and Hedo will play just fine against Odom/Kobe defense. Kobe's a good defender, but he's giving up 4 inches to either jump shooter. The magic do a good job of ball movement, and Kobe tends to roam, so I'm expecting several catch and shoot opportunities that Kobe will be just a split second too late to do anything about.
And so far Phil says he'll start Bynum/Gasol and have Lamer come off the bench. I think that limits his effectiveness. But even if he does switch back to starter, I still think Hedo will be able to take him off the pick and roll. Hedo runs that like a guard.
Gasol will give Dwight a rough go though. I mean, big baby Davis had some good games from mid-range against Dwight.
can't stand Kobe or the Lakers - that being said , like Joe stated, the matchups favor the Lakers . . . I have to hope that Odom and Bynum continue to show disappearing acts half the time . . .
For the good of the country I think Prest should resign as GM an become some sort of counter-intelligence czar. "Thunder very interested in Ben Gordon..Rip Hamilton..McGee..Harden..Curry..Mullins..Thabeet..Rubio" The only thing that is clear at this point is that Presti has been extremely thorough, and that nobody has a real clue as to his intentions. On second thought, the country can take care of itself- let's keep that guy.
HATE THE LAKERS!!!!!
I am obviously pulling for Orlando, but they have their work cut out for them. The problem I see is that the matchups aren't in Orlando's favor like they were in the Cleveland series. In the Cleveland series you had 6/3" Delonte West trying to guard Turk; you saw how that went. And you had Varejao and Joe Smith and Ben Wallace trying to guard Rashard; good luck with that.
In the Lakers series, Odom can handle either of those two if he's healthy, and Kobe is also a good defender. Bynum and Gasol I think will be in hot water with Howard, but by the same token, Howard is going to have to guard Gasol out a lot farther than he had to with big Z. Big Z couldn't hit a shot, so Howard didn't have to respect it. Also, Pietrus was huge in the last series as a change of pace guy who plays good defense, much like Posey did for Boston last year. He can be covered I think pretty effectively by Ariza..
I just see the matchups a bit in the favor of LA.
I'm pulling for the Magic in this series. I dig Kobe, but I like Dwight more. I like that inside/outside game that the Magic run. And, I always cheer for the underdog when I don't have a vested interest in either team.
@Ty
I'm hoping they don't pick that high. Also, it's a year away. None of those guys may look any good next year.
Who would the Thunder take at PF if they had a chance to next year
Demarcus Cousins
Derrick Favors
Greg Monroe
Ed Davis
a great article
http://www.welcometoloudcity.com/2009/6/2/896596/2009-nba-draft-coverage-hello-my
Back to reality, I thought it was interesting that you said the Magic aren't known for defense, Royce. Orlando had, possession for possession, the best defense in the entire league this year. Not a big deal, just funny to me.
I think two is a little misleading also. It's not that you don't need a great PG, it's that you do need a PG who can play within the offense well. This is really the reason that I still want Westbrook at the 2 sans an incredible increase in his game-management abilities. He'll get better, for sure, but our biggest issue on offense is the lack of cohesion. Westbrook's game, in general, is spontaneous and unpredictable. It is beautiful chaos. Unfortunately, that's kind of the opposite of what you want in your lead guard, the guy who is supposed to set up an offensive approach for the whole team.
@Pennington
You can shake hands and be upset - the two things are NOT mutually exclusive . . .
Durant was probably upset at his team for quitting - I'm sure his anger was not directed at the other team . . .
I am glad Durant was like that too, that's what we need . . .
still, there is NO excuse for not being a good sport . . .
@David
I would feel bad if seattle had supported the team or even cared when it was happening. Don't blame Clay blame the people and legislature of Seattle for failing to support the team. Be mad at the previous ownership for selling the team, or David Stern for allowing the relocation. But don't be mad at Clay and especially don't be mad at the people who live in Oklahoma they did nothing but support The Hornets and want a team. We didnt want to steal your team, we just wanted a team. Besides its been said before Seattle hates Oklahoma for having the team, we get it. Now please shut up. I have better things to do than to listen about how much you hate Clay and Aubrey two men whom I know personally and who I respect.
I Like LA in 6 too.
My favorite moment of the year, we were down 20 or something with a few minutes left in the 4th. Everybody had given up. Jet Zero, Uncle Jeff, Thabo, and Nenad were all just going through the motions. KD was the only player that showed any heart. At one point we turned the ball over and from my seats in sec. 117 I could hear Durant scream a four letter word. He was upset that we lost and that there was nothing he could do. It was at that point that I knew he was bound for superstardom. I want the studs to be upset when they fail. I would much rather have Durant blow off handshakes and a post game press conference following a playoff exit, than smile and shake our opponents hands as we have 20 win seasons. I want to know that he cares as much about the team as we do. I really think he does.