Smart an upgrade over Bradley

[KG Living Legend]

 This is a post on Celtics Subreddit. The stats are interesting.

 Our starting lineup last year, IT/AB/Crowder/Horford/Amir, was pretty [dang] good. We went 26-10 when it was intact, which is 59-win pace over a full season. Over 432 minutes of play, that lineup outscored opponents by +6.2 points per 100 possessions, compared to our full season rating of +2.8.

However, our best lineup, by a significant margin, was when we replaced AB with Smart. It had an outstanding +17.1 net rating. It wasn't a small sample size either: at 192 minutes, it was our second most used lineup.

According to both BPM and RPM, our starters slightly underperformed. BPM expected them to have a rating of +9.8, while RPM thought they should be +7.7. This can probably be accounted for by playing against the other teams' starters. Conversely, the Smart-ers were much greater than the sum of their parts, with expected ratings of 'only' +10.0 (BPM) and +10.1 (RPM).

So what does this mean for this season? AB and Amir, alas, are gone, but we have some pretty handy reinforcements. Based on last year's numbers, a possible starting lineup of IT/Hayward/Crowder/Morris/Horford would be expected to rate somewhere between +9.3 and +11.5. Swapping Baynes with Morris as a token starter limits the range somewhat (+9.2 to 10.4) but would solidify our bench.

Jaylen Brown, unfortunately, rates horribly by last year's advanced stats, even though he performed admirably with the starters when AB was injured. As such, an IT/Jaylen/Hayward/Crowder/Horford unit is expected to produce a 'mere' +6.0 to +8.6.

Finally, the advanced stats are unanimous on what should be our closing lineup. Replace Jaylen with Smart, and the expected rating jumps to +9.5 (RPM) and +11.7 (BPM).

p.s. this all assumes that everyone performs at the same level as last season. plug in the infamous CARMELO projections instead, and the results are a bit more bland:

IT/Smart/Hayward/Crowder/Horford: +7.8
IT/Hayward/Crowder/Morris/Horford: +6.7
IT/Hayward/Crowder/Horford/Baynes: +5.5
IT/Jaylen/Hayward/Crowder/Horford: +4.9

[Roy H.]

Our second most used lineup only played 192 minutes together? Can that be right?

[KG Living Legend]

 It makes sense. A little less than half of the starters minutes.

[Ilikesports17]

Our second most used lineup only played 192 minutes together? Can that be right?

I'm always shocked to see how little minutes your most common lineups get together.

[CelticsElite]

This is why smart will start at the 2. Not brown.. Also brown plays best with Tatum who comes off the bench

I'm brown will do the occasional start depending on opponent but I see smart moving up

[Greyman]

Despite the stats, which are affected by a number of things like the mentioned opposition lineups, I don't see Smart as an upgrade but possibly the best option to start in AB's place. With Brown looking to get more minutes and probably improving on last season Smart isn't a shoe in but I think he will be given first shot at that position, it is a good time for him. The pressure to score will be less with Hayward present and this could help him.

[mctyson]

Our second most used lineup only played 192 minutes together? Can that be right?

Think about how many 5-man combinations are used during a season and it will make sense. 

[mctyson]

However, our best lineup, by a significant margin, was when we replaced AB with Smart. It had an outstanding +17.1 net rating. It wasn't a small sample size either: at 192 minutes, it was our second most used lineup.

To follow-up my point above, 5-man lineups are not really worth looking at because there are so many combinations, you are bound to find one that confirms your belief (and you can find one that contradicts it). 

For example, if you take the IT/Smart/Horford/Crowder core and sub-in Kelly O for Amir through 100 minutes that 5-man had a -13 net rating, whereas if you then sub-in Bradley for Smart, that new 5-man of IT/Bradley/Horford/Crowder/Olynyk is +13.  Does that make Smart worse than Bradley?

I think the 3-man combinations are a bit more reliable in that the minutes played sample size is larger and you can see who Brad considers to be the core:  Crowder, Horford, IT.  It will be interesting to see how Hayward fits.

[Surferdad]

Despite the stats, which are affected by a number of things like the mentioned opposition lineups, I don't see Smart as an upgrade but possibly the best option to start in AB's place. With Brown looking to get more minutes and probably improving on last season Smart isn't a shoe in but I think he will be given first shot at that position, it is a good time for him. The pressure to score will be less with Hayward present and this could help him.

This, on two counts: (1)  There are lots of variables unaccounted for with this kind of purely statistical argument but the most significant is probably opposition lineups.  Smart was first off the bench so probably played more minutes against opponents bench players, at least in the 1st and 2nd quarters. (2) That said, Smart is indeed the best option to start in AB's place but not for the reason claimed by KGLL, but more what Greyman said.

I think we will also see a lot of IT/Hayward/Crowder/Morris/Horford, especially if Smart is not fast out of the gate this season.

SIDE NOTE: It seems almost every thread lately winds up to be a discussion about Marcus Smart. His performance this season is key for both the team success as well as his own future.

[loco_91]

Overall, I agree with the "Smart>Bradley" sentiment. However, I'm still not convinced that we can move Smart to the 1st unit without sacrificing too much in the 2nd unit. When AB was injured last season, Stevens chose to plug Brown in as starter, due to Smart's key role in the 2nd unit offense. In the playoffs, he went so far as to start Green for Amir, even though his preferred small ball lineup has Smart in for Amir.

I expect we'll start IT/Hayward/Crowder/Morris/Horford.

[BlackCeltic]

Overall, I agree with the "Smart>Bradley" sentiment. However, I'm still not convinced that we can move Smart to the 1st unit without sacrificing too much in the 2nd unit. When AB was injured last season, Stevens chose to plug Brown in as starter, due to Smart's key role in the 2nd unit offense. In the playoffs, he went so far as to start Green for Amir, even though his preferred small ball lineup has Smart in for Amir.

I expect we'll start IT/Hayward/Crowder/Morris/Horford.

I agree. Bringing Smart off the bench also keeps his numbers modest. The better he does the more money he will seek next offseason. Look for Smart to be a little more selfish this year.

[SHAQATTACK]

omg yes.

[More Banners]

This is why smart will start at the 2. Not brown.. Also brown plays best with Tatum who comes off the bench

I'm brown will do the occasional start depending on opponent but I see smart moving up

Yeah. I'm expecting Smart to slide into AB's spot. Hope he has a jumper.

Anyway, I want to see Brown develop, so would prefer that, but really, its Smart. Some have suggested Crowder, but I just don't see that pairing - Hayward/Crowder on the wing - being quite as good.

[billysan]

I won't argue with the stats, I'm sure they are accurate. That will be an impressive closing unit.

I do maintain that it is likely that Morris, Crowder and Smart come off the bench together as the second unit. Starters are IT, Brown,  Hayward, Horford, and Baynes.

The reason for this is the starters need to be our best offensive unit. Bringing those three in against second unit players will shut down opposing teams defensively. Combine them with any 2 starters except Baynes and we still have good offense IMO.

I just don't think your closing 5 should always be your starting 5 best players. 
.
Balanced lineups are more important.

[droopdog7]

Is smart an upgrade over AB.  Not to my eyes.  Should smart or brown start?  Neither is my answer.  Start Hayward and Crowder at the two and three.