First of all, you're a loyal member of the community and have already read the Downbeat for today -- so you know what Ross Siler tweeted. If you don't know, Ross Siler was one of the beat writers for the Jazz and while he has moved onto another career path he still maintains a great interest and relationship with the Jazz and we, the Jazz fans. You can tweet at him here. Because I can't pass up a good PNG file, here it is again:
It makes sense, especially for a team in need of some cohesion, to go with guys who have actually played with one another before this season. Raja Bell knows Andrei Kirilenko, but back when they were running up and down the court they were playing with Carlos Arroyo at point - not Deron Williams. This is just one example of unfamiliarity in our lineups, but it's a good one. Because I'm a huge nerd, I immediately went to 82games.com to look for instances of this Deron / CJ / AK / 'Sap / Memo line up in the last few years. I did not find any. That said, I did find 15 line ups from 2006-2007 till today that featured four of the five guys. Deron is in all 15. C.J. Miles is in 8. Andrei and Paul are both in 12. And bringing up the rear, Mehmet Okur is in 13. (Nice number) If you add it all up (and I did - 60 out of a total 75 matches, or 80% accuracy), I think we can look at how some of those lineups have fared. From this we may be able to extrapolate how well this potential line-up could be for us in the short term.
Click below to read on, loyal blog reader . . .
Initial points and Methodology
Before I go too deep into this, let me state that I'm already working on a big article on line-ups. A lot of the data used here will be used again there. Furthermore, a lot of the terms used here will definitely be used in the next big post. In order to evaluate line-ups (5 man units) there exist a number of categories. Some are obvious, like +/- (did this lineup outscore who they played against) -- while others flow naturally (like win%, which uses W/L record of that line-up each time it's used, as a reflection of that time frames particular +/-). Offensive categories are points per possession (Off. PPP), and eFG% (which I use frequently on this site). Defensive categories are opponents points per possession (Def. PPP), and Opponents eFG% (Opp. eFG%). Last, there are two more categories: rebound percentage (how many available boards does this unit grab), and turn over differential (like a +/- for turn overs, but represented in a % number). I think these are fair and valid categories to use.
In the case of evaluating this potential line-up (again, there were a total of 101 different lineups that I found on 82games.com over the past 5 years and none of them were of the five guys Ross mentioned above) I had to collect all of the line-ups that featured four of these five players. There were 15 such line ups. As much of the cumulative data as possible was averaged by the number of lineups. I did not weight the different lineups by another other factor, like minutes played first. (I tried, but couldn't get it to work)
How valid are these numbers? Well, these numbers are influenced by the fact that a) they do not represent data from an actual Deron / CJ / Andrei / Paul / Memo lineup; and b) are influenced by data infection from the 15 remaining line-up spots that we are overlooking. This remainder exists as: Ronnie Brewer (4), Kyle Korver (4), Carlos Boozer (3), Derek Fisher (1), Matt Harpring (1), Wesley Matthews (1), and Al Jefferson (1). I can't really do anything about that. Thus, these figures are approximate and hypothetical. Furthermore, we don't know the true, Gestalt affect these guys will have on one another until we see the unit in action.
Back to the numbers, some categories were easier to deal with - like cumulative factors like +/- and minutes played. Wins and losses were also easy to figure out (and as an extension - win%). Offensive and Defensive PPP were hard because 82games uses a different format for games from 2006-2007 and earlier. I just had to throw out the data from those three lineups for these two categories. I took the cowards way out and just averaged the remaining 12 lineups from 2007-2008 till 2010-2011. I did the same thing for eFG% and Opp. eFG% (but used all 15 line-ups worth of data). I wish I had the actual FGM-FGA / 3PTM-3PTA for these groups so the data would be more precise. I do not, so this is what you get.
This hypothetical Siler lineup looks okay on paper. The numbers also show this. They would have had an average minutes played of 90.58 (100+ mins together seems like something we can bank on, and not write off as a fluke). During this time period (the full 1358.7 minutes played over 15 units), they went 124-101, for a Win% of 55.1%. That's not a world beater - but it's a start. That record would be tied for #35 best winning % out of 101 total lineups over the last five seasons. This group had a cumulative +/- of +203, which divided by 15 gives you a +/- of 13.5.
On the offensive side of the ball this unit averaged 1.19 points per possession (# t-17) and an eFG% of 51.3% (#54). On defense, this unit allowed the opponent to have 1.05 points per possession (#28) and and opponents eFG% of 47.8% (#52). Round out the eight categories are rebound % and turn over %. This team would get 73.8% of the available rebounds (#40) and be a +1.2% better than the bad guys when it comes to keeping a hold of the ball (#47).
Sure, by these numbers alone they don't seem like the best possible line-up we've ever had. And true, this line up doesn't crack the Top 5 in any of these eight categories against all the lineups we've had this season alone -- except, well, one important one. Rebounds. The Top 5 units this year for rebound % are:
- Williams, Price, Miles, MIllsap, Jefferson (60.0 rebound%)
- Watson, Price, Miles, Elson, Fesenko (59.5%)
- Williams, Miles, Kirilenko, Millsap, Jefferson (57.0%)
- Watson, Price, Hayward, Elson, Fesenko (56.5%)
- and Watson, Price, Miles, Kirilenko, Fesenko (53.0%)
This Ross Siler unit would gobble up 73.8% (theoretically), which is better than all of these listed units, and also better than the average over 101 different units as well. If you wanted to see the averages for the 101 total units, here they are:
- Win%: 51.4%
- +/-: +8.2
- Off: 1.12 PPP
- Def: 1.06 PPP
- Opp eFG%:47.7%
- Reb%: 70.6%
- TO%: 1.2%
Ross' lineup would is 55.1 win%, +/- +13.5, Off 1.19 PPP, eFG% 51.3%, Def 1.05 PPP, Opp eFG% 47.8%, Reb% 73.8%, TO% 1.2%. This lineup is better than the average everywhere except for virtual ties in turn overs and opponents eFG%. Frankly, at this stage I'd settle for 'average' of what we've had the last 5 years.
Beyond the Numbers
True, these numbers will carry this conversation only so far. These guys haven't produced these numbers, they are extrapolations from 15 units worth of data over 1358.7 total regular season minutes. I am a huge believer in a unit's ability to augment (or diminish) certain players own abilities. Clearly Ronnie Price's feistiness is assisted by Earl Watson being willing to pressure the defense in ways Jason Hart never would. Similarly so, we know that Deron loves playing with C.J. Miles, and vice versa. They look for each other on the break, and more than anything else -- run the break on broken plays. If you've been following the Jazz for a while you know that Andrei and Memo are even tighter on and off the court. Back when Boozer first came here and was injured we say a lot of Andrei and Memo playing off of one another. Memo's best season (statistically) came with Boozer out of the way, and Andrei passing him the ball. In both of these cases I see a chemistry benefit to this line up. Millsap is loved by all, really. He's the anti-Boozer (or at least was) in terms of hard work and willingness to do the dirty work. If the offensive focus on this new team is put on Deron, CJ and Memo I think Millsap can go back towards being a great complimentary player (instead of a struggling primary player).
Defensively I'd take this group as well. Okur is great at showing off of pick and rolls (much better than Boozer or Jefferson for sure), and Miles and AK are long enough to close out on shooters and make a difference. Millsap hasn't been defending that well this season, but I think the team defensive schemes may work a little better with this group than a number of guys who don't know what they are doing (as a unit).
The largest concern is that the data that all of this is culled from comes from back when Memo was a 16 and 8 guy. He's not the same guy right now, and he's not going to be a consistent 15 and 6 guy for another few weeks (months). Jefferson for all his faults is trying hard and improving. He just doesn't know the offense to where he can be killing it right now. Which is the lesser of the two evils, Jefferson's noob status, or Okur's rust?
The next concern would be relegating Millsap to a smaller role. Yes, he would be a smaller role player in this group, but at the same time, he wouldn't be competing for post up position against Jefferson. Memo will be spotting up while Andrei and CJ are slashers.
CJ and Andrei are quite complimentary wings. CJ is a hot shooter who can get on a roll, and Andrei can be surprisingly effective when he's aggressive on offense. CJ can score from everywhere and Andrei has a knack for getting to the line. That said, they are both kind of iffy handling the ball, and both kind of streaky with their scoring efficiency. Miles is under rated as a defender and AK still has something left in the tank - but it's hard to see Jerry going with these two for long stretches when he seems to have been playing either guy at SF, replacing one guy for the other / not playing them together. It remains to be seen that Jerry may have also made some alluring promises vis a vis playing time / starting security to Raja Bell this off season. (Not verified)
I think it's worth trying out. I was surprised to see that out of 101 line ups from the last 5 years he hadn't seen this one before. Obviously, Memo of today isn't the Memo we all love. I think that's the biggest problem right now. Playing guys who don't know the plays / don't know each other is playing line-ups at the expense of winning games. Lots of vocal Jazz fans have strong feelings about playing certain guys at the expense of winning games. I think we need to recognize the outcome (losing) can happen even if you play vets. Sub-optimal lineups have been killing us this season. This lineup may not be the most optimal, but I believe that having guys on the same page will provide a sense of purpose and unit cohesion that we've so far lacked this year (outside of The Swarm). Be on the look out for more line-up analysis as well.