Why UConn misses the three-guard lineup

In the game thread of the Tennessee debacle Justin brought up an interesting question: how much of the team's struggles the past two games can be realistically attributed to the lack of Ryan Boatright? After all Boatright's pace adjusted stats (via Kenpom, subs. required) aren't much of an improvement over Shabazz Napier's, so it's easy to view the pining for Boatright as just another case of UConn fans over-valuing the temporarily unavailable. (Remember, prior to his arrival last season, half The Boneyard talked about Enosch Wolf the way Samuel L. Jackson talked about Winston Wolf.)

On my site last Friday I noted that part of the reason Boatright is important is that his value as a third guard and second ball handler transcends his statistical value as a player. Look at the affect that the three-guard lineup has on Jeremy Lamb.

Jeremy_lamb_statistics_by_numbers

There is nothing in Boatright's stats that suggest that Lamb's spike in output is due to Boatright in particular. Rather, as Shabazz Napier noted after the Cincinnati game, it's due to having an extra ball-handler in the game. The question then becomes, would UConn have won the past two games with the three-guard lineup available?Moreover for a lineup which, as Jim Calhoun has noted, has difficulty rebounding the ball, do the increases in offensive efficiency outweigh what the team concedes in size, rebounding ability and (presumably) defense?

I started tackling the issue in my last post here, but I've since re-run the data with an eye for more specific defensive metrics. The results were pretty surprising. First a look at the offensive side of things:

Tub_-_3-guard_offense

As expected, the 3-guard lineup is the most efficient offensively. Although the unit concedes offensive rebounding ability, they shoot and score more efficiently, make more threes, and turn the ball over less. It is interesting that there is such a drop-off in the performance of the Napier/Lamb backcourt during the games in which Boatright was eligible. I'm not sure why that is, but among the possible factors is the possibility that competition was stronger during that sample, so I don't find it all that surprising or odd.

What is surprising, are the defensive numbers:

Tub_-_3-guard_defense

In addition to being more efficient offensively, the 3-guard lineup posts the best numbers nearly across the board on defense. The 3-guard unit holds opponents to a worse shooting percentage, forces more turnovers and grabs an equivalent number of defensive rebounds. They also foul slightly more often, but not so much more that this offsets their gains in other areas.

It appears that what the 3-guard backcourt concedes in size, they've made up for in aggressive, on-ball defense. Napier has alluded to this on more than one occasion noting that he and Lamb were able to play more aggressively on defense knowing that Boatright was available off the bench (or alongside them on the court). A look at the shot-distribution chart shows that they've been most effective in limiting opponents' jump shots, even if they've given up slightly more at the rim.

Tub_-_shot_selection

So this re-foucuses our original question: can we quantify the value of the 3-guard lineup? This is different than quantifying Boatright's value as I'm not going to look at lineups in which Boatright replaced Lamb or Napier.

Before I get into the calculations here, let me acknowledge that this is something I'd refer to as counter-factual, statistical, alchemy--which is an incredibly pretentious way of saying "hypothetical bullshit." Making this kind of estimate requires controlling too many uncontrollable factors to call it definitive. But it works in the vacuum of theory, so what the hell...

For the purposes of this exercise I'm going to assume that the dip in the output of the Napier/Lamb backcourts during the games Boatright was eligible is due exclusively to superior competition. Based on this assumption I'm going to compare the outputs from the 3-guard lineup to the outputs from the Napier/Lamb backcourt during the games Boatright was eligible. (Even if the dip was due to other factors, I think this gives us the most accurate reference point because it controls for competition and number of games). Therefore I'm going to use the Napier/Lamb numbers from the games played with Boatright eligible as a stand-in for Napier/Lamb backcourts in the Cincinnati and Tennessee games.

During the games Boatright was eligible the Napier/Lamb backcourt scored 1.03 points per possession while the 3-guard lineup scored an average of 1.16 points per possession. Over the same sample the Napier/Lamb backcourt played roughly 34.1% of possessions, while the 3-guard lineup played roughly 41.54% of possessions. Using the percentages of possessions played and UConn's average pace of 65.4 possessions per game (via Kenpom, again), we can extrapolate that the 3-guard lineup is worth an additional 3.65 points/game on offense.

Offensively the boost from the 3-guard lineup already exceeds the margin of defeat, but when we add the defensive component it gets even better. The 3-guard lineup played 41.34% of defensive possessions and gave up .951 points per possession. The Napier/Lamb backcourt played 33.29% of defensive possessions and gave up .994 points per possession. Using the 65.4 possessions per game number again, we can extrapolate that the 3-guard lineup would've prevented 1.16 points in an average game.

In total, swapping the 3-guard lineup in for the Napier/Lamb backcourt on 41.34% of possessions is worth a net of 4.81 points per game. This exceeds the 3-point margin in both losses since Boatright's been suspended (or has been awaiting suspension, as the case may be).

As I said before, this proves nothing definitively. It fails to account for the other lineup combinations on the floor, counter-moves by opposing coaches, or UConn's ability to make stat-padding, desperation 3's and come back during garbage time. It also doesn't absolve the team of some all-around terrible performances. As Calhoun said after the Cincinnati game, while Boatright would've made a difference he could have been out due to injury or foul trouble. While the Cincinnati game was easy to brush off as an odd matchup (their 4-guard lineup exploited UConn's lack of backcourt depth), the loss to an inferior Tennessee team (even without Boatright) is tougher to swallow.

This does show how Boatright's value exceeds his statistical output though, because the lineup flexibility is a huge piece of his value. It also helps explain some of the offensive ineptitude we've seen the past two games. Considering all of Calhoun's National Championship teams have had two proficient ball handlers (Moore/El-Amin, T. Brown/Gordon, Walker/Napier), it shouldn't come as a surprise that the second PG is such an important piece of the puzzle. If nothing else, it's more fuel for the fire when you check out the second part of Joe Nocera's New York Times piece on Boatright tomorrow.

Check out Carl Jackson at his blog UConn By the Numbers or his twitter @Uconnbynumbers.

SB Nation Featured Video
X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join The UConn Blog

You must be a member of The UConn Blog to participate.

We have our own Community Guidelines at The UConn Blog. You should read them.

Join The UConn Blog

You must be a member of The UConn Blog to participate.

We have our own Community Guidelines at The UConn Blog. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9341_tracker