Here’s a basic rundown of how the Wins Produced metric works (see Calculating Wins Produced for the full rundown)
[Player Box Score Stats] + [Team Defense] – [Team Help] – [League Average for their Position]
- Player Box Score Stats is just the weighted box score numbers for any player.
- The League Average compares the player against all other players at the same position.
- Team Defense looks at how good the player’s team’s defense was and assigns the player a percentage of that based on how many minutes they played.
- Team Help looks at how good the player’s team was at helping (assisting and blocking) and either rewards the player if they played on a bad team (the player was better than they looked) or ding them if they play on a good team (the player was worse than they looked).
Most people are ok with Player Box Score Stats and League Average. The Team Defense and Team Help is where some claim that we wave our hands to make everything right. Today I’ll be reviewing Team Defense and Team Help with the 2011 Season to help alleviate this fear.
Team Defense relies on five separate team stats.
- Three Pointers made by the opponent
- Two Pointers made by the opponent
- Opponent Turnovers that weren’t a result of a steal.
- Team Turnovers – these are turnovers not directly attributed to a given player (e.g. shot clock violation, or in other words, turnovers that are not steals)
- Team Rebounds — these are rebounds not directly attributed to a given player that resulted in a change of possession.
As we can see, Team Defense primarily checks how good the team is at impacting the opponent’s made shots and forcing turnovers. Let’s see the impact of the Team Defense on each team.
|Portland Trail Blazers||19805||44.9||0.109||6.9||0.017|
|New Jersey Nets||20005||24.4||0.059||1.6||0.004|
|New Orleans Hornets||19830||43.3||0.105||1.5||0.004|
|Los Angeles Clippers||19855||32.6||0.079||1.1||0.003|
|Los Angeles Lakers||19830||57.5||0.139||-0.4||-0.001|
|Oklahoma City Thunder||20005||51.8||0.124||-3.8||-0.009|
|San Antonio Spurs||19780||56.3||0.137||-4.3||-0.01|
|New York Knicks||19780||42.9||0.104||-5.2||-0.013|
|Golden State Warriors||19830||34.1||0.082||-5.5||-0.013|
Table 1: Team Defenses for 2010-2011 Season
Team Defense can certainly make a difference. A large part of Chicago’s surge in the East can be attributed to it. Some of the downright terrible teams in Cleveland, Minnesota and Toronto had atrocious defenses. What is important to note is that all of the defensive wins fall within 11 Wins Produced, and for a majority of teams (19) it’s less than five wins. The responsibility for a good or poor defense is also shared across many players. Kevin Love for instance earned over 25 wins this season. He was credited with -0.7 wins as a result of his defense (he played roughly 13% of the available minutes). If somehow we blamed him for all of the Wolves terrible defense he’d still be worth over 18 games.
Team help relies on two stats
- Team Assists
- Team Blocks
This actually works the opposite way as Team Defense. If you’re on a team that passes a lot and has good shot blockers your numbers are slightly inflated. Players lucky enough to play with the likes of Steve Nash and Jason Kidd fall under this category. Here’s a rundown of how teams are affected by team help numbers.
|Golden State Warriors||19830||34.1||0.082||-2.0||-0.005|
|New York Knicks||19780||42.9||0.104||-1.5||-0.004|
|Los Angeles Lakers||19830||57.5||0.139||-1.3||-0.003|
|San Antonio Spurs||19780||56.3||0.137||-1.2||-0.003|
|Los Angeles Clippers||19855||32.6||0.079||-1.1||-0.003|
|Oklahoma City Thunder||20005||51.8||0.124||0.7||0.002|
|Portland Trail Blazers||19805||44.9||0.109||1.3||0.003|
|New Jersey Nets||20005||24.4||0.059||1.6||0.004|
|New Orleans Hornets||19830||43.3||0.105||2.4||0.006|
Table 2: Team Help for 2010-2011
Not a bunch of surprises here. If your team had good playmakers and defensive stoppers the team numbers take a slight hit. It was surprising to see Utah at the top of this list but they did have Deron Williams for part of the season and were in the top five both in assists and blocks. We can also notice that this adjustment is much smaller than team defense, and all teams pretty much stay within 5 wins.
Putting It All Together
|Portland Trail Blazers||19805||44.9||0.109||8.2||0.020|
|New Orleans Hornets||19830||43.3||0.105||3.9||0.009|
|New Jersey Nets||20005||24.4||0.059||3.2||0.008|
|Las Angeles Clippers||19855||32.6||0.079||0.0||0.000|
|Los Angeles Lakers||19830||57.5||0.139||-1.7||-0.004|
|Oklahoma City Thunder||20005||51.8||0.124||-3.1||-0.007|
|San Antonio Spurs||19780||56.3||0.137||-5.5||-0.013|
|New York Knicks||19780||42.9||0.104||-6.7||-0.016|
|Golden State Warriors||19830||34.1||0.082||-7.5||-0.018|
Table 3: Total Team Adjustments for 2010-2011
When we put the numbers together only one team manages to get 10 wins from the team adjustment. Turns out Milwaukee was great on defense and terrible on help, and as a result had the biggest swing in team adjustment. Here are some fun teams to note.
- Chicago was in the top five for most helpful team and team defense. This reinforces the notion that Chicago was a really good team and it wasn’t all Derrick Rose.
- Orlando and Milwaukee were in the top five for defense and the bottom 5 for team help.
- Phoenix was in the bottom five for defense and top five for help.
When looking at the team adjustments for wins produced it is clear that they do matter. Many of the fears people have (adjusting for pace on teams like the Suns, accounting for defense) are taken care of with these. To be fair these adjustments aren’t perfect, but what’s key to notice is that at the player level these adjustments are not large. In the worst case scenario — such as Phoenix and Milwaukee — even a player getting more than 2,000 minutes won’t notice a boost or hit of much more than 1.0 Wins Produced. To be a truly great player (such as Kevin Love) or a truly terrible player (such as Andrea Bargnani) it requires a lot of heavy lifting on the player’s part themselves (and no team adjustment will fix that). And this is why the correlation between the adjusted and unadjusted production numbers (as noted in this forum since 2006) is around 0.99.
Hope that answered some questions and I’m sure there may be more. Feel free to leave them in the comments (just make sure they are polite and we can continue explaining how things work).