Monta and Steph: Homies for Life?

Yesterday, I posted a fanshot linking to an article by Sebastian Pruiti in his NBA Playbook blog where he attempted to bust the myth that Monta Ellis and Stephen Curry can't play together. This of course stems from Monta Ellis himself, as he famously said at media day last year:

-Q: Were you OK with them drafting Stephen Curry? He’s pretty similar to you…

-ELLIS: I mean, I don’t have a problem. Whatever decision they make upstairs, that’s their decision. They’re doing it for a reason.

I’m going to wish Curry the best in training camp, get better as his career goes on. We’re not worrying about it. We’re teammates. We’re not out here fighting for a position, because that’s not the case.

-Q: Can you see yourself playing with Curry in the backcourt?

-ELLIS: I can’t answer that. Us together? No.

-Q: Why not?

-ELLIS: Can’t. We just can’t.

-Q: Too small? Too similar?

-ELLIS: Just can’t.

-Q: The Warriors say you can.

-ELLIS: They say we can? Yeah. If they say it. But we can’t.

-Q: You wouldn’t want to give it a shot?

-ELLIS: I just want to win. That’s… not going to win that way.

-Q: Why not?

-ELLIS: It’s different when you’re trying to compare me and Stephen, when you’re trying to go back to when me and BD were playing, it’s a different situation.

You’ve got a nine-year veteran who’s been in the game, who understands the game, knows how to play the game, and he’s a big body…

You can’t put two small guys out there and try to play the 1 and the 2 when you’ve got big 2 guards in the league. You just can’t do it.

OK, yes, we’re going to move up and down fast, but eventually the game is going to slow down.

You can’t do it

(Source) .

Pruiti made several excellent points about how Curry and Ellis played last year, and how they play together. He provides some video evidence, some excellent analysis of the X's and O's of Curry and Monta playing together, and comes to the same conclusion that many of Warriors fans who watched Curry and Monta play together at length last year would come to, that it's plausible that they will struggle defensively, but they should play well together offensively.

The Curry/Ellis backcourt is actually something that I have been looking at indepth recently. So I thought I'd add my own analysis to this. I have some issues with a couple of statements that Pruiti makes, first there's this:

The offensive end is where this duo can really excel.  Both players can shoot and be a threat on the offensive end, and that makes it tough for defense.  In fact, if you visit BasketballValue.com, you can see that just about every lineup that includes Ellis and Curry had an Offensive Rating of over 100 (the only one that didn’t featured Vladimir Radmanovic and Mikki Moore, which probably hurt on the offensive end).

To understand why this isa dubious statement, you need to have an understanding of what Offensive Rating means: points per 100 possessions. The league average offensive rating last year was 107.6. The Suns had the best offense in the league last year, with a 115.4 offensive rating. The Warriors on the season had an offensive rating of 107.8. The Nets were last in the league with an offensive rating of 100.3. So when you say that almost every lineup with Ellis and Curry had offensive rating of over 100, you are saying that almost every lineup had an offensive rating better than the Nets. Or, to put it another way, you are saying almost every lineup was no worse than 7 points worse than the average Warriors lineup. Basically, saying that they had offensive ratings higher than 100 means that they were better than one of the worst teams in NBA history.

A better way to look at how Ellis and Curry played together is to look at how the Warriors did with and without each player on the court:

The first thing to notice is that the when Ellis was on the court, the offensive was pretty much the same, with or without Curry. But when Curry played without Ellis, the offensive flourished. I haven't been able to get individual totals when for each of these on/off states, I would imagine that Ellis's usage rate hovered around 30% whether or not Curry was on the floor and that Curry's usage rate dipped slightly below 22% when Ellis was on the floor, but his, and everyone else's, went up slightly when he was off the court. Basically, I imagine that a general observation that pretty much everyone made during the season, that the Warriors move the ball better when Ellis isn't around, holds true.

Pruiti noted that the biggest weakness of this backcourt would be defense, and he was definitely right about that. Here's how the Warriors performed defensively with and without Curry and Ellis:

It's worth noting, once again, that league average for defensive rating is the same as offensive rating, 107.6. The Warriors, as a team, were second worse in the league, 111.4. The Bucks and Bobcats were the best defensive teams last season with ratings of 103.2 and 103.3, respectively. The only team worse than the Warriors was found north of the border, where the Raptors had a pathetic defensive rating of 113.2.

Looking at Curry and Ellis together, defense was definitely a problem. When they played together, opponents put up a ridiculous 113.84 offensive rating. The Warriors defense improved somewhat when only one of them were on the floor together (although, let's face facts: it was never pretty on defense last year). This confirms Pruiti's (and I'm guessing most Warriors fans') suspicions that the two are going to have trouble guarding larger guard combos.

When we put all of this together, we get the following +/- ratings for Curry and Ellis:

The first thing that jumps off the page is that with Curry and without Ellis, the Warriors had a +5.5 rating. For a team that was overall -2.13, that's a pretty amazing result. The other thing to notice is that the Warriors were by far at their worst when Curry and Ellis played together.

Based on this evidence, it's not hard to see that Monta Ellis was right. Ellis and Curry didn't play well together, and the Warriors did not play winning basketball with them playing together. Of course, this method of analyzing how they played together has it's flaws. Plus/Minus stats are noisy, and I am ignoring the other three or four players on the floor for the Warriors, and ignoring the quality of competition they played against, and the sample sizes aren't that large. However, even taking those issues into account, I think the conclusion is still valid.
One thought that I had, and I'm sure many other Warrior's fans are having, is that Curry's season really picked up in January. Certainly, the Ellis and Curry started playing better together once Curry got comfortable with the pro game and started playing better? Here are the numbers from January 1st on:

Well, they performed a little bit better from January on, but not much. So there goes that theory.

Why Hopefully None of This Matters
Last year's season was pretty much doomed from that fateful media day. This season, there already is a different vibe with this team, as evidenced by what Ellis is saying now:

When I made that comment, I wasn’t knocking him, I wasn’t bashing him. It was based off of what me and the Warriors were going through at the time. It had nothing to do with him. Like I told him this summer, we can do it. We did it last year. I see now that I have to play the two (shooting guard). I have to play the bigger guards, and I’m cool with that.

(source)

Monta has gone from "Just can't" to "We can do it." I think this is one of the biggest developments in an offseason of big developments. If Ellis is willing to change his game, and play more of a team game, it can only help the Warriors. There's still the problem of defending the other team's guards, and that is going to be very important to this team's success. With the additions of Lee and Amudson, and the return of Biedrins, the Warriors should be able to secure the rebound after pretty much every missed shot Curry and Ellis can force.

I've been a big fan of pretty much every personnel move the Warriors have made this off-season (only exception is the Udoh pick, but that's for another day), but so much of this season's success hinges on Ellis and Curry playing well together. They have the potential to be a very potent back court. Let's hope they live up to that potential.

**UPDATE**

I made some Venn Diagrams to show how Curry and Ellis played with the Warriors other wings from last season:

Curry, Ellis, and Watson:

Curry, Ellis, and Morrow:

Curry, Ellis, and Maggette:

Curry, Ellis, and Williams:

Some things to note:

• When Curry and Watson played together, good things happened. Watson is highly underrated. He was an above average shooter and passer, a passable defender, and, probably most importantly for this team, he didn't turn the ball over that often.
• Whenever Reggie Williams stepped on the floor, scoring went up. For both team.
• Monta's best pairing was with Williams, but no group with Monta had a positive rating.
• Curry was at his best when he played with Watson, and without Maggette or Ellis.

**ANOTHER UPDATE**

I wanted to get some idea of the quality of competition for the earlier Curry/Ellis splits, so I decided to take a page out of jae's playbook and look at the number of starters they faced (weighted for time):

So, going off this, we can see that the Warriors definitely faced the toughest competition when Ellis and Curry played together. This could partially explain why the Warriors do worse when they play together.