Which lineups are working best (and worst) so far for the Knicks?

Going into this season, the general consensus was that the Knicks would not be a very good team, but I don't think many people were expecting New York to be 3-10 after 13 games. Of course, there are reasons for the slow start: new schemes on offense and defense, a first-time head coach, the injury to Jose Calderon, etc. However, it really boils down to the fact that Derek Fisher has a very limited amount of talent to work with.

As such, if he wants to maximize the potential of this team, Fisher needs to identify player groupings that will be successful and stick with them. Further, he needs to avoid using lineups that will amount to a dumpster fire. Luckily, there's a very easy way for us to take a look at which player groupings have worked best so far, as Basketball-Reference's lineups page allows you to see just that.

Before we get started, a few notes on what I'm about to show you:

• All of the stats listed below are Net Per 100 Possessions. These numbers signify the difference between what the Knicks have done and what their opponents have done, so if the number is positive the Knicks have been better than their opponents, and if it's negative the Knicks have been worse. The "Per 100 Possessions" part makes it easier to compare the lineups to each other, since it's very unlikely that any lineup would have had the exact same number of possessions as another one.
• I did the research on this prior to Wednesday's game in Minnesota, so these stats only reflect anything that happened up to and including the game in Milwaukee on Tuesday.
• It's important to note that it's still very early in the season and therefore the sample sizes here are small. The reason I point that out is because the fact that these sample sizes are small means that the numbers can change pretty quickly. It's possible that everything I say here will be completely useless by the end of the season (or even in a month). That being said, the lineups that came out looking the best made sense based on what I've seen in games.

Okay! First, I wanted to check out the five-man lineups that Derek Fisher trots out most often. I decided, very arbitrarily, to only look at lineups that had played at least 12 minutes together (again, through the first 12 games). This left me with seven different combinations. The table below lists those seven, with the lineups sorted by Net Points Per 100 Possessions.

I would say that these results are not that surprising. As you'll see when I look at three-man groups, Carmelo Anthony-Iman Shumpert-Tim Hardaway Jr. has been an enormously successful threesome ("ha-HAAAA!" - Clyde), so it makes sense that the two lineups performing the best would involve those three. Of course, it's seriously unlikely that either lineup continues to outscore opponents by 40 or 30 points per 100 possessions, but the fact remains that those two lineups have been impressive.

It may help to take a peek at the other statistics, to determine how fluky the numbers are. The Melo-Timmy-Shump-JR Smith-Amar'e Stoudemire group is out-shooting opponents by 23.3% overall and 6.9% from three while averaging almost exactly the same amount of assists as opponents, but they're getting out-rebounded by 5.3 per 100 possessions. The only number that seems completely out of place here is the +23.3 FG%, which is definitely not sustainable. However, that FG% may also actually boost the lineup's rebounding numbers, since they have many more opportunities for defensive boards than their opponents --meaning that, as the FG% comes back down to Earth, the rebounding may get even worse. I would expect this lineup's net points to drop fairly quickly.

On the other hand, the Melo-Timmy-Shump-Shane Larkin-Samuel Dalembert lineup should continue to beat down opponents, especially once Jose Calderon replaces Larkin. The fact that this lineup is registering 8.3 more assists per 100 possessions than opponents is unsurprising, and the excellent rebounding numbers from this lineup make sense based on the +19.8 FG%. Even if the rebounding and shooting regress, I can see the assists and three-point shooting staying about where they are, which bodes well for this group.

Looking at the other end of the spectrum, things don't look good for the Melo-Shump-Larkin-Dalembert-Jason Smith quintet. Even though the overall FG% should probably get better, the three-point shooting may get worse because the only real three-point shooters in this lineup are Melo and Shump. Also, I can't imagine the rebounding getting much better here.

Next, let's take a look at three-man groups. I made the cutoff here 84 minutes (again, totally arbitrary) and came up with 12 groups. These are also sorted by Net Points Per 100 Possessions.

Only three of the 12 groups are outscoring opponents, which I guess shouldn't be much of a shock considering the Knicks' record. The numbers for Melo-Shump-Hardaway actually look fairly sustainable. It's a bit of a shock that they're being out-shot from behind the arc, but that shouldn't last long, especially since Fisher likes to play the three of them alongside JR and a center a good amount of the time.

A group I'd expect to get better points-wise as the season progresses is Melo-Shump-Dalembert. Even without Calderon in the lineup, that group is out-shooting opponents, both overall and from three, as well as getting more assists. However, since the Knicks lose a lot and will probably continue to do so, a group that plays this high a percentage of minutes may have no shot at ever out-scoring opponents consistently.

Finally, let's take a look at some two-man pairings. The arbitrary cutoff here was 144 minutes, and I again sorted by Net Points Per 100 Possessions.

Among these nine pairings, only Melo-STAT has a positive number in the net points column, and even that rating is barely above zero. I don't want to look too much into any specific pairing, but an interesting pattern here is that each of the nine pairs shoots better overall, even though only two of the nine (Melo-Larkin and Dalembert-Larkin) shoot more than one percent better than opponents from behind the arc. However, given New York's historically abysmal three-point defense, maybe this shouldn't come as a shock. Also, I found it odd at first that Tim Hardaway, Jr. didn't show up in any of these pairings...but then I remembered that there are some games in which he barely plays, so perhaps it makes sense.

Really, what this data tells us is the Knicks stink and probably will continue to do so. And hey, maybe that's by design. But if it's not, then I guess my advice for Derek Fisher would be this: play Carmelo Anthony, Iman Shumpert, and Tim Hardaway, Jr. together as often as you can, because they score a lot more points than their opponents do. Oh, and you should probably never start Jason Smith again. Maybe don't play him at all, really. He sucks.