Best 5-man units

[Wham with the Right Hand]

As we get to the 60-game mark I thought I would look at some of the Cavs’ most commonly used lineups and see which ones have been the most effective. (I’m a big believer in chemistry and putting complementary skills on the floor).

I used the website cleaningtheglass.com since it eliminates garbage time and end-of-the-quarter heaves.

Starting with the lineups that include Garland, Mitchell (DG/DM), Mobley, and Allen, since it’s clear this is our core:

Point differential is per 100 possessions, the # of possessions is in parenthesis.

1. DG/DM/Okoro/Mobley/Allen +10.4 (445 possessions)

This is the current starting lineup. It’s point differential is in the 74th percentile of all 5-man units with over 100 possessions.

2. DG/DM/LeVert/Mobley/Allen +6.1 (339)

Subbing in LeVert for Okoro weakens the unit by 4.3 points, but overall the lineup still is in the 58th percentile, or slightly above average.

3. DG/DM/Stevens/Mobley/Allen -3.1 (301)

Inserting Stevens at small forward results in a lineup that gets outscored despite having our four best players on the floor. This just hasn’t worked.

4. DG/DM/Wade/Mobley/Allen +30.3 (33)

Only 33 possessions with this unit which is too small a sample for reliable information, but I’d like to see more. I think potentially this could be our best unit.

5. DG/DM/Osman/Mobley/Allen +50.5 (48).

Another small sample but awesome results.

Now for lineups without Mitchell:

1. DG/LeVert/Okoro/Mobley/Allen +6.0 (190)

Not bad when LeVert replaces Mitchell at the 2. It’s almost exactly the same as when LeVert replaces Okoro at the 3; the point differential drops from about 10 to about 6.

2. DG/LeVert/Stevens/Mobley/Allen +13.7 (94).

This lineup with Stevens and LeVert has been very good in about four quarters worth of action.

Lineups without Garland:

1. DM/LeVert/Okoro/Mobley/Allen: -25.6 (116)
2. DM/LeVert/Wade/Mobley/Allen: +24.5 (105)

Almost the same number of possessions but Wade makes the unit 50 points better than Okoro. I’m sure with more possessions those numbers will converge.

As for the individual players, their on/off numbers line up like this:

Cedi +6.2
Allen +5.7
Wade +5.7
Garland +5.0
Rubio +2.3
Mitchell +1.2
Neto +0.4
Okoro -1.7
Mobley -2.7
Love -4.3
LeVert -7.4
Lopez -13.8

I keep expecting Cedi to fall back but we’re 3/4ths of the way through the season and he still has the best on/off number, although he normally is on the floor when opponents are resting some of their starters. However, Cedi has been in the negative each of the last four years, including a -8.6 last year, so his number has improved a whopping 14.8 points since a year ago.

Last year he was the worst on the team; even Rondo, Denzel Valentine, Kevin Pangos and Ed Davis were better.

 

[foucault87]

As we get to the 60-game mark I thought I would look at some of the Cavs’ most commonly used lineups and see which ones have been the most effective. (I’m a big believer in chemistry and putting complementary skills on the floor).

I used the website cleaningtheglass.com since it eliminates garbage time and end-of-the-quarter heaves.

Starting with the lineups that include Garland, Mitchell (DG/DM), Mobley, and Allen, since it’s clear this is our core:

Point differential is per 100 possessions, the # of possessions is in parenthesis.

1. DG/DM/Okoro/Mobley/Allen +10.4 (445 possessions)

This is the current starting lineup. It’s point differential is in the 74th percentile of all 5-man units with over 100 possessions.

2. DG/DM/LeVert/Mobley/Allen +6.1 (339)

Subbing in LeVert for Okoro weakens the unit by 4.3 points, but overall the lineup still is in the 58th percentile, or slightly above average.

3. DG/DM/Stevens/Mobley/Allen -3.1 (301)

Inserting Stevens at small forward results in a lineup that gets outscored despite having our four best players on the floor. This just hasn’t worked.

4. DG/DM/Wade/Mobley/Allen +30.3 (33)

Only 33 possessions with this unit which is too small a sample for reliable information, but I’d like to see more. I think potentially this could be our best unit.

5. DG/DM/Osman/Mobley/Allen +50.5 (48).

Another small sample but awesome results.

Now for lineups without Mitchell:

1. DG/LeVert/Okoro/Mobley/Allen +6.0 (190)

Not bad when LeVert replaces Mitchell at the 2. It’s almost exactly the same as when LeVert replaces Okoro at the 3; the point differential drops from about 10 to about 6.

2. DG/LeVert/Stevens/Mobley/Allen +13.7 (94).

This lineup with Stevens and LeVert has been very good in about four quarters worth of action.

Lineups without Garland:

1. DM/LeVert/Okoro/Mobley/Allen: -25.6 (116)
2. DM/LeVert/Wade/Mobley/Allen: +24.5 (105)

Almost the same number of possessions but Wade makes the unit 50 points better than Okoro. I’m sure with more possessions those numbers will converge.

As for the individual players, their on/off numbers line up like this:

Cedi +6.2
Allen +5.7
Wade +5.7
Garland +5.0
Rubio +2.3
Mitchell +1.2
Neto +0.4
Okoro -1.7
Mobley -2.7
Love -4.3
LeVert -7.4
Lopez -13.8

I keep expecting Cedi to fall back but we’re 3/4ths of the way through the season and he still has the best on/off number, although he normally is on the floor when opponents are resting some of their starters. However, Cedi has been in the negative each of the last four years, including a -8.6 last year, so his number has improved a whopping 14.8 points since a year ago.

Last year he was the worst on the team; even Rondo, Denzel Valentine, Kevin Pangos and Ed Davis were better.

Thanks for the great post as usual Wham. I still cannot make heads or tails of how the +/- is being distributed from the players to the lineups sometimes. Especially with Mobley who visually is clearly a better player than last year and at worst our 4th best guy...but a -2.7 over the season AND in all our top 5 man units. I bought that it was noise when we were 15-20 games into the season, but at this point it is looking like a real number and one I haven't heard a great explanation for. Same applies to a lesser extent to Okoro and Caris

 

[Wham with the Right Hand]

Thanks for the great post as usual Wham. I still cannot make heads or tails of how the +/- is being distributed from the players to the lineups sometimes. Especially with Mobley who visually is clearly a better player than last year and at worst our 4th best guy...but a -2.7 over the season AND in all our top 5 man units. I bought that it was noise when we were 15-20 games into the season, but at this point it is looking like a real number and one I haven't heard a great explanation for. Same applies to a lesser extent to Okoro and Caris

Yeah, I'm confused as well. NBA.com has Mobley with a net rating of +5.3. I sent a question to CTG to ask why the discrepancy. Here are the NBA numbers:

Players Advanced | Stats | NBA.com

A table featuring advanced information for each player in the league based on selected filters.

www.nba.com

Both websites have Mobley 9th on the team in on/off, but the numbers are drastically different. They are in good agreement as to the order. NBA has the top five as Allen, Wade, Osman, Garland, and Mitchell. CTG is slightly different: Wade, Allen, Osman, Garland, and Mitchell. I didn't count Rubio since he has barely played.

 

[Smooth]

Great work as always, @Wham with the Right Hand..

Cedi and Wade have pretty consistently been near the top of +/- rankings for the Cavs this year.. Love was at one point too before his shooting fell off a cliff..

With respect to Cedi and Wade’s production, think their +/- speaks to the importance and value of shooting around the core four.. Guys who can provide spacing are incredibly valuable to the starters and overall well being of the team..

Think it’s also why you see guys more towards the bottom of the list being worse shooters than those near the top..

If you’re able to replace Cedi with a more consistent presence or even just add more consistent shooters to the rotation, you’d see the value they bring to the team..

This isn’t to say the +/- is strictly swayed one way or another by shooting but the impact it has on this teams success I think is somewhat seen in these number..

 

[Wham with the Right Hand]

What is interesting about the NBA +/- list is the defensive ratings. Rubio is at 100.6 (in just 14 games) while Wade is right behind at 101.0. After those two there is a big fall-off to the next best defender, which is Cedi at 107.0. In fact, the rest of the regular rotation players are all in the 107-108 range. It's just Wade and Rubio who are well above that level and that might be because both of them have played limited minutes due to injury.

 

[Wham with the Right Hand]

If you’re able to replace Cedi with a more consistent presence or even just add more consistent shooters to the rotation, you’d see the value they bring to the team..

That might be the reason (or one of them) as to why they signed Danny Green. Some nights Cedi just can't find the range. Usually it's obvious from his first couple of shots. Green gives Bickerstaff another option.