Survey Project: Wins Added by Season

[ElGee]

"How much better does he make a team? How many wins is that guy worth?"

That's the goal of this survey project -- for real gm posters to answer this one question (Guidelines below). Please read the project information below if you'd like to participate...

Guidelines:

Spoiler:

Take each year of a player's career and estimate how many wins he would add to a random, 41-win team, based on his play that year. You've likely done this before, many times, even without realizing it. Simply take the player in question and add him to an average team.

--It's the summer of 2002, and Player X signs as a free agent with the Orlando Magic. The 2002-03 Magic were an average team. Knowing everything you know about what happened in 2003 (Injuries, player changes, etc.) how many wins will the 2003 Magic win now that they've added Player X?

--It's the summer of 2006, and Player X signs as a free agent with the Los Angeles Clippers. The 2006-07 Clippers were an average team. Knowing everything you know about what happened in 2006 (Injuries, player changes, etc.) how many wins will the 2006 Clippers win now that they've added Player X?

Two Rules:

(1) Base your projection on how well the player played at the END OF THE SEASON (i.e. "playoff time"). If he's injured, gimpy, tired, etc. then assume he would have played like that throughout the year on a random, average team.

(2) Do not adjust for games missed. Take the version of the player at the end of the season and assume he plays 82 games that way on a random, average team.

Do this for a few different average teams -- some average teams need defensive anchors. Some need shooting. Some need players who can create their own shot. Then take the average number of predicted wins and write it down. That's the survey! The first few answers may take some thinking through, but it will quickly become intuitive...

(In essence, this means you need to predict how much the player would improve more than one team. Here are some average teams this century http://www.basketball-reference.com/pla ... =wins)

NOTE: Negative values are possible due to replacing a neutral player.

Wins or SRS and a Sample answer

Spoiler:

example: In 2007, how many wins did Shotblocking Sam add to a random, average team? Here's one way to go about this...

In 2007, Sam is added to the Magic -- a good defensive team with a young Dwight Howard and Tony Battie. On the other hand, they don't have many players who can create their own offense or for others. They have decent shooters. Adding player Sam's defense and rebounding can upgrade Battie, and upgrade the interior when Howard's on the bench.

Let's also add 2007 Sam to the Lakers. Here he can provide a nice upgrade at center, providing defense and rebounding to a team that struggles in both areas.

**Run this thought experiment a few more times, then take the average, then write it down next to "2007" on Sam's ballot**

Note: Some people are more comfortable looking at team quality in terms of point differential (SRS). Answers can be posted in Wins OR SRS -- it does not matter. Some people may also want to post a range (eg 54-56 wins), this is fine as well, but for the survey results I will use the average. Posting the average or the range is fine.

*Sam's Sample Ballot*
2005 38 W
2006 41
2007 45
2008 47
2009 47
2010 44

or
2005 -1
2006 0
2007 2
2008 3
2009 3
2010 2.5

Participation:

Spoiler:

The project is open to anyone -- all that is asked is that participants enter a ballot for each player. It will be difficult to use your ballot if you only submit one for one or two players.

Yes, in this project it's OK if you enter a ballot without discussion or explanation. If you want to participate, just indicate so in this thread or PM me, and I'll add your name to the list.

Furthermore, if you want to catalogue data about the player in his thread then feel free to do so without signing up for the survey. This includes posting statistics, anecdotes, articles, etc.

NB: PLEASE do not argue with, criticize or comment in any negative way on other people's survey answers. Feel free to ask people about their reasoning, but this not a mind-changing, discussion-driven project. It's a survey and archival project, and if you want to debate people's assessments, well, I think that's what the rest of the threads on the board are for.

Structure:

Spoiler:

All player threads will be open for an indefinite period of time. Every few days a new one will be added. The starting scope of the survey is the latest realgm top 25 list, in order. After that, any wild-card players can be surveyed based on what is desired.

Results:

Spoiler:

**Will be updated based on survey responses...

Below are the links to each thread:

Real GM Survey: Value by Season threads
1) Michael Jordan
2) Bill Russell
3) Kareem Abdul-Jabbar
4) Magic Johnson
5) Wilt Chamberlain
6) Larry Bird
7) Shaquille O'Neal
8) Tim Duncan
9) Hakeem Olajuwon
10) Kobe Bryant
11) Julius Erving
12) Karl Malone
13) Kevin Garnett
14) Oscar Robertson
15) Jerry West
16) Moses Malone
17) Dirk Nowitzki
18) LeBron James
19) Bob Pettit
20) Charles Barkley
21) David Robinson
22) Dwyane Wade
23) Walt Frazier
24) Steve Nash
25) Scottie Pippen

List of participants

Spoiler:

Ardee
Colts18
Doctor MJ
Gregoire
JordanBulls
Narigo
Penbeast0
Sideshow Bob
ThatsWhatIShved
thebottomline

[JordansBulls]

I'll participate if ok when I get back in town.

[ceiling raiser]

I'd love to participate in this project, though I'm going to have spend some time looking over my data first.

[ElGee]

I'd love to participate in this project, though I'm going to have spend some time looking over my data first.

That's fine -- don't feel like you have to rush into answers. I'll be leaving the threads open indefinitely and won't be doing any tabulation of answers right away, so take your time.

[Dr Positivity]

I will participate

[SideshowBob]

Thread's for:

Kobe Bryant

Lebron James

Kevin Garnett

[lorak]

How many SRS points is one win worth?

[ElGee]

How many SRS points is one win worth?

The formula is non-linear, based on Oliver's formula in BOP:

Wins SRS
8 -15
9 -14
11 -13
12 -12
14 -11
15 -10
17 -9
19 -8
22 -7
24 -6
27 -5
29 -4
32 -3
35 -2
38 -1
41 0
44 1
47 2
50 3
53 4
55 5
58 6
60 7
63 8
65 9
67 10
68 11
70 12
71 13
73 14
74 15

In other words, if you're an average team and you find a way to regularly outscore other people by 3 pts/game (eg 93 Shaq shot at 93% from the FT line that year), on average the team will win 9 more games. If 93 Shaq shot 25% from the line, the 93 Magic would be expected to win 32 games (assuming they were a 0 SRS team...they were actually +1.4).

For those unfamiliar with SRS, it's used over wins because it's abetter predictor. Wins has some "noise" from close games, which almost no teams in history consistently outperform their expected results based on points (SRS) in the long haul, thus SRS is more accurate (no noise).

[ElGee]

Thread's for:

Kobe Bryant

Lebron James

Kevin Garnett

Thanks Sideshow -- I think I'm going to reboot these just for consistency. When I do, I'll import the current responses.

[Gregoire]

Great project! And where other players threads? I wanna see it for Shaq,Wilt ect

[ElGee]

Great project! And where other players threads? I wanna see it for Shaq,Wilt ect

Just opened Bill Russell and will launch a new one every few days or so. In the meantime, feel free to prepare answers for the other players...

[NO-KG-AI]

I've been reading these (and got your PM), it's just so hard for me to formulate answers for this that I feel confident about. "average" team is hard, especially when an average team can be built and tailored so differently. I'll try to get some answers in, but I'll be reading, and probably commenting for sure.

[ElGee]

I've been reading these (and got your PM), it's just so hard for me to formulate answers for this that I feel confident about. "average" team is hard, especially when an average team can be built and tailored so differently. I'll try to get some answers in, but I'll be reading, and probably commenting for sure.

I agree the first time you conceptualize this it's a bit challenging...but in essence it's really an attempt to quantify an opinion you already have. Obviously, as you said, the challenge is in mentally surveying the different shapes of teams. As mentioned in the OP, I think the easiest starting point is looking at a few .500 teams and specifically adding the player in question to their roster -- if this were some FA signing in the offseason, how much would you now like the offense and defense?

I'd also add not to fret about confidence -- it's really an attempt at quantifying an opinion you "already have."

[colts18]

I've been reading these (and got your PM), it's just so hard for me to formulate answers for this that I feel confident about. "average" team is hard, especially when an average team can be built and tailored so differently. I'll try to get some answers in, but I'll be reading, and probably commenting for sure.

I agree the first time you conceptualize this it's a bit challenging...but in essence it's really an attempt to quantify an opinion you already have. Obviously, as you said, the challenge is in mentally surveying the different shapes of teams. As mentioned in the OP, I think the easiest starting point is looking at a few .500 teams and specifically adding the player in question to their roster -- if this were some FA signing in the offseason, how much would you now like the offense and defense?

I'd also add not to fret about confidence -- it's really an attempt at quantifying an opinion you "already have."

Right now the average teams in the league are the Celtics, Lakers, Jazz, Hawks, and Mavs. I'm confident that adding LeBron to any of those teams with the exception of the Hawks could lead to a title. The Jazz compare favorably to LeBron's Cavs teams. They have a nice big man rotation of Millsap/Jefferson/Kanter/Favors. They have Mo Williams who played for LeBron. They have 3 40%+ 3P shooters at 3PA/game (Williams, Foye, and Hayward). He could be competitive with that.

The Mavs and Lakers are obviously title teams with a LeBron or Wade type of player too.

[JordansBulls]

I've been reading these (and got your PM), it's just so hard for me to formulate answers for this that I feel confident about. "average" team is hard, especially when an average team can be built and tailored so differently. I'll try to get some answers in, but I'll be reading, and probably commenting for sure.

I agree the first time you conceptualize this it's a bit challenging...but in essence it's really an attempt to quantify an opinion you already have. Obviously, as you said, the challenge is in mentally surveying the different shapes of teams. As mentioned in the OP, I think the easiest starting point is looking at a few .500 teams and specifically adding the player in question to their roster -- if this were some FA signing in the offseason, how much would you now like the offense and defense?

I'd also add not to fret about confidence -- it's really an attempt at quantifying an opinion you "already have."

Right now the average teams in the league are the Celtics, Lakers, Jazz, Hawks, and Mavs. I'm confident that adding LeBron to any of those teams with the exception of the Hawks could lead to a title. The Jazz compare favorably to LeBron's Cavs teams. They have a nice big man rotation of Millsap/Jefferson/Kanter/Favors. They have Mo Williams who played for LeBron. They have 3 40%+ 3P shooters at 3PA/game (Williams, Foye, and Hayward). He could be competitive with that.

The Mavs and Lakers are obviously title teams with a LeBron or Wade type of player too.

Pretty much agree with your here, but also the average team such as the Lakers and Celtics and Mavs are teams that do have stars in place but just not playing to there potential. Reminds me of the 1990 Hawks who had Dominique and Moses and were .500 team and the 1998 Rockets with Hakeem, Barkley, Drexler and were a .500 team. So adding any top 3 player in the league to those type of teams will turn them into championship favorites instantly.

[colts18]

Pretty much agree with your here, but also the average team such as the Lakers and Celtics and Mavs are teams that do have stars in place but just not playing to there potential. Reminds me of the 1990 Hawks who had Dominique and Moses and were .500 team and the 1998 Rockets with Hakeem, Barkley, Drexler and were a .500 team. So adding any top 3 player in the league to those type of teams will turn them into championship favorites instantly.

The Bulls are average this year too. Just adding Rose would make them a title contender. Add a Wade or leBron and they are the favorites.

[SideshowBob]

Ahh Wilt and Bird next. Wilt's going to be a bit controversial. How will we go about rating him, particularly in his uber-volume years? Based on what we know who he is or based on what we think he could have been? Since we're trying to encapsulate "goodness," I feel inclined to go with the former (and I'm particularly high on 64/67/68).

[GSP]

Timmy shall be very interesting IMO prolly the most consistent player since Kaj.

[ardee]

Timmy shall be very interesting IMO prolly the most consistent player since Kaj.

MJ wants a word with you.

[excision]

(1) Base your projection on how well the player played at the END OF THE SEASON. If he's injured, gimpy, tired, etc. then assume he would have played like that throughout the year on a random, average team.

What's the reason for this?