Applying A 2013 NYU Study On "Roster Synergy" To Determine Ideal Terps Lineups, Substitution Patterns, Recruiting

I just came across this fascinating 2013 NYU-Polytech Study, published int he International Journal of Computer Science in Sport. To say the least, I am a pig in shit.

First, the authors propose a way in which you can calculate a team's "synergy" (or how well the parts compliment each other):

For each NBA team, we formed lineups using the top five players in terms of plays played in our data sample. We calculated their ratings individually and as the 5-player lineup. For a given lineup of players x1, x2, x3, x4 and x5, define PORP(x1,x2,x3,x4,x5) to be the estimated point differential between a game played by this team of players against a lineup of replacement players ("RP"). We then define synergies as the difference of the sum-of-the-parts from the team total

The idea is relatively simple. You look at how many points a lineup would be expected to score compared to a lineup of replacement level players (based on the individual ratings of the players in the lineup) and then you see how many points above replacement they actually scored. The difference, is the "synergistic effect." If the sum of a team's parts is somehow greater than the individual parts--they have great synergy. If the opposite is true, they have poor synergy.

What the researchers noted is that, in the 2006-2007 NBA season, the difference between the team with the most synergy (The Orlando Magic) and the least synergy (The Minnesota Timberwolves) was worth about 2 points per game. Using widely accepted metrics, they concluded that a team who maximizes their synergy will win roughly 6 additional games per year, while teams who minimize their synergy will lose about 6 additional games per year. If you translate that to a college basketball schedule, that would amount to about 2.5-2.7 wins or losses.

Next, the researchers tried to answer the question of what maximizes or minimizes synergy. Analyzing the data from the 2006-2007 season they found:

Our framework predicts that skills that affect rare events (e.g. steals, offensive rebounds) will have positive synergies, while skills that contribute to common events (e.g. defensive rebounds) will have negative synergies....

Elsewhere in the study they found that there was basically a highly negative synergy between players who shoot a lot (i.e. high "usage rate"). Think: pairing up Russell Westbrook and Kevin Durant. There's only one ball and so you are essentially wasting K.D.'s talent when Russel Westbrook isos and takes a shot, and visa versa.

So there you have it. Putting together guys who share skills in rare events creates the best synergy. That isn't to say that it creates the best team--a team with K.D. and Westbrook is still going to own a well synergized team with inferior talent--it just means you are maximizing your talent level.

This is an interesting study to me because, looking at the Terps, you see a team with basically a bunch of players who are pretty similar in terms of how they are rated by WS/40. (note: WS/40= "win shares per 40" and it rates how many wins a player contributes to for each 40 minutes of playing time) Thus, creating positive synergies and minimizing negative synergies seems more important on a team without any true stars.

What is the most synergistic "featured 5" the Terps can field?

While the Terps lack any truly elite players based on individual WS/40 there are four guys who I believe have to be in the "featured lineup" just based on WS/40 (i.e. they are just clearly our best four players):

PG- Seth Allen: He's got a WS/40 of .150 and he's not even up to speed yet. Roddy Peters and Nick Faust, the only other options here, both have significantly lower WS/40 (.122 and .053 respectively).

G/F: Dez Wells: He's got the 2nd highest WS/40 on the team and does a little bit of everything well. Also I think most would agree based on the eye test that he is our best player.

F: Jake Layman: He's got the highest WS/40 on the team. And by a significant margin (.163 compared to Dez's .151).

F/C: Charles Mitchell. He's got a WS/40 of .142. Shaq has a WS/40 of only .09 and Jonathan Graham is hovering around .105. He's also the best defensive rebounder on the team (DR% of 22%) and offensive rebounder (OR 17%).

That leaves Evan Smotrycz, Nick Faust, Jonathan Graham, Shaq Cleare, and Roddy Peters competing for the other spot in our best lineup. So who should it be? All of these players are about the same in WS/40 (hovering in the 1.0-1.3 range except for Roddy who is significantly lower at .053). So synergies are going to be really important in determining who we should play.

The first thing you need to look at is what are the potential areas for positive and negative synergies.

Potential Positive Synergies

Steals: Jake, Dez, Chuck, and Seth all force between 1.3 to 1.6 steals per 100 possessions which is significantly above average. Adding an elite steals man to this would create a positive synergy.

Of the remaining players: Nick and Roddy create really good synergies here. (2.8 stls/100 and 2.5 stls/100 respectively). Evan creates a really good synergy. (1.8 stls/100). Shaq and Graham are non-factors (1.0 stls/100 and Graham has yet to force a steal).

Score: Nick 5; Roddy 4; Evan 3; Shaq 2; Graham 1

Offensive rebounds: Chuck is a beast on the offensive glass. Adding additional offensive rebounding creates a positive synergy.

Graham wins out here (9.8 percent of offensive boards) but Shaq and Evan are not slouches (7.8 and 7.7 percent of offensive boards respectively). Nick and Roddy are lesser factors on the offensive glass (4.8 and 2.2 percent respectively).

Score: Graham 5; Shaq 4; Evan 3; Nick 2; Roddy 1

Blocks: In the group of Chuck, Dez, Seth and Jake, there are no elite shot blockers. But they aren't slouches as a group. Chuck, Dez and Jake avg 2.8, 3.4 and 3.6 blocks per 100 possessions respectively. Seth is less of a factor with less than 2 blocks per 100 possessions. So adding a good shot blocker would still have a positive synergy because the rest of the team provides good depth to go along with that.

Graham wins out here as well. He's actually a pretty damn good shot blocker, blocking 7.3 percent. Shaq is a distant second, but his 3.9 blocked shots per 100 possessions would also have a positive synergy and give the lineup some good shot-blocking depth. All in all this is a really strong synergy for Graham. His good shot blocking should be boasted by really good shot-blocking depth. Nick, Evan and Roddy are non-factors.

Score: Graham 5; Shaq 4; Evan 3; Nick 2; Roddy 1

Overall Positive synergy scores: Graham 11; Shaq 10; Evan 9; Nick 9; Roddy 6.

Negative Synergies of Note

Evan takes a HUGE hit here in my mind. His two best traits are his shooting (although he's in an extended slump there) and his defensive rebounding. Thus there will be a negative synergy with Layman's shooting and Seth's shooting, and a negative synergy with Chuck's defensive rebounding.

Nick, Shaq, and Graham don't seem to have any of their main traits have significant negative synergies, with the exception that Graham does take hit because his defensive rebounding ability is somewhat redundant, given Chuck's ability in that area.

My best 5's:

My best 5 are Seth, Dez, Jake, Shaq/Graham, and Mitchell

Another acceptable "featured" lineup to me if the match-ups favor going small is Seth, Nick, Dez, Jake, Mitchell.

Good Substitution Patterns:

Never play Evan, Seth, and Jake at the same time (really negative synergies b/e their shooting attributes). All three would like to be high usage rate shooters. But really only one, maybe two, of them can do that at a time without wasting some of their value.

Try not to play Evan and Chuck together. (negative synergies between their team best defensive rebounding, more positive synergies created by Shaq and Graham).

Playing Roddy and Seth at the same time can actually work. (creates one less shooter for Seth to be free to be a high usage rate offensive player, and there is a really good synergy between Roddy's/Seth's steal percent and the rest of the lineup's).

The bigs (Shaq, Chuck and Graham) should be played together more often. (creates positive synergies in blocks, offensive rebounding that Evan doesn't bring to the table)

When Evan is out on the floor, that would be a good time to play Shaq who is not a good defensive rebounder (minimize the negative synergy between Evan's good defensive rebounding and Chuck/Graham's good defensive rebounding).

All in all this analysis would create less opportunities for Evan to play, with many of his minutes going to Shaq/Graham.

Future Recruiting (Reed a really good pick up; Turge is recruiting too many shooters):

The fact that Graham creates the best synergies with our best four players is troubling. He's not a very good player. We could upgrade that by adding another big who is a good offensive re-bounder and an elite shot blocker. Reed seems to fit that bill. Dodd seems capable of developing into that. Shaq's development is also a wildcard in this mix. Thus, we are on the right track to creating good synergies with Chuck, Dez, Jake, and Seth if we want the 5th player in 2014 to be a big.

However, the study I suggested really calls into question Turgeon's strategy of recruiting so many guys who are shooters. Having a bunch of highly rated offensive shooters isn't necessarily bad. However it does make for a team whose parts are less than the whole and creates an overall negative synergy per the NYU study.

If you are going to recruit a bunch of guys who excel at one particular trait, you want that trait to be a rare one: like steals or offensive rebounding. Given that we have Chuck (great offensive rebounder) and have a bunch of guys who can steal the ball--recruiting players of similar caliber to Wiley, and Nickens, but who are ball hawks on D and who will crash the glass on offense would be a better strategy. In any case if we were going to take Wiley (shooter) it made little sense to add Nickens (also a shooter). Especially when one considers that our lineup currently features Seth (shooter-high volume); Jake (shooter--needs to be at a higher volume for us) and Evan (who, unfortunately, seems to want to be a high volume shooter). We are simply going to be wasting a lot of our shooting talent.

The old adage: you can never have too much shooters isn't necessarily false. If you have a team full of Ray Allen caliber shooters at all 5 positions you are obviously going to win the NCAA title. But given that we live in the real world, a world of opportunity costs and in which the Terps are limited in terms of the caliber of recruit we typically pull in--synergy maximization should play a big part in deciding who we recruit. Thus, adding a lot of high 3* to high 4* shooters at the expense of adding similar caliber players with positive synergies is not a great strategy.

Anything deemed inappropriate will be deleted by an admin or moderator with the power to do so. The views of the above FanPost do not represent the beliefs of Testudo Times or Testudo Times' authors, nor are they the work of them.

Log In Sign Up

Log In Sign Up

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior users will need to choose a permanent username, along with a new password.

Your username will be used to login to SB Nation going forward.

I already have a Vox Media account!

Verify Vox Media account

Please login to your Vox Media account. This account will be linked to your previously existing Eater account.

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior MT authors will need to choose a new username and password.

Your username will be used to login to SB Nation going forward.

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 Testudo Times

You must be a member of Testudo Times to participate.

We have our own Community Guidelines at Testudo Times. You should read them.

Join Testudo Times

You must be a member of Testudo Times to participate.

We have our own Community Guidelines at Testudo Times. You should read them.




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.