Stock Watch 2015-16 Meta-Analysis
Jun 20, 2016 18:56:28 GMT -5
billy, Paul Pierce, and 4 more like this
Post by Allan Houston on Jun 20, 2016 18:56:28 GMT -5
This data isn't perfect, but it should be close. There was some wonkiness over the course of the season, mostly with Brown Cobb trying to override other GMs' votes. But for the most part, things were in order, and if not I did my best to fix them.
This season, there were 152 players put up for stock watch, 151 of which had their rating changed (Sorry Nurkic; but thanks for inspiring us to leave polls open longer)
The following GMs voted the most:
Thanks everyone for their votes. Those of you who neglected to vote, you're encouraged to do so next season! More votes = more accurate results.
The average change was +1.9. It's a little worrying that there's rating inflation, but I think one mitigating factor is that there are more likely to be rating increases than decreases. There were 87 rating increases this season, compared to only 62 decreases. Nurkic, again, didn't get enough votes to change. Brandon Bass and Al Horford were voted to remain at the same rating.
Why are there more likely to be increases than decreases? For one, it's always beneficial to stock watch your own guys who are due increases. For decreases, GMs need to target other teams, and that takes a bit more effort, and comes with fear of reprisal. Furthermore, underrated players who are unrostered are picked up and stock watched. Overrated players who are unrostered remain unrostered, with no one bothering to stock watch them.
Also, it's telling that the median change was +1. Since the mean is larger, that means that increases were larger than decreases. Part of that is because there were some <45 players who were increases to rotational level guys, whereas decreases would never go below 45.
Which players had the largest changes in a single stock watch?
Which GMs deviated from the final result the most? Who were the most negative and who the most positive? Using mean signed deviation, the most generous, by far, voter was Doc Rivers. Granted, he only voted 3 times, so hasn't gotten a feel for it yet. That said, he overrated players, relative to the final votes, by and average of 8 1/3 points.
Following Doc is Tony Stark, who in his 2 votes overrated players by an average of 4.5 points. Steven Rogers was +3 average above the result in his 6 votes. Next is Kareem, averaging +2.6 in his 13 votes. Then we have Hanamichi, RIP, +2.05 in 17 votes. And finally, Steve Rogers, +2 in 6 votes.
The differentials on the other side, of those who were much less generous in their votes, are a bit less drastic. That's mostly because they tended to vote more often. It seems like new guys come in and overrate players, whereas the vets are more cynical. As we can see, the most votes among the top five most generous voters were Hanamichi's 17. You'll notice a difference with this next group.
The most negative voter was Milt Newton, who voted on average 1.6 points less than the result, having voted 20 times. Ignoring Ghost Billy King, bc I don't know who those 7 votes really belong to, I come in next at -1.5. To be honest, I thought I'd top this list, often noticing that I voted a player the lowest. But I voted every time for a total of 152 votes, so outliers don't affect much. In third is Theodore Duncan, who in 25 votes averaged -1.16. Next is Keith Van Horn, -1 in 8 votes. Finally, rounding out the top (or should I say bottom) five is Logan Hough, -.66 in 80 votes.
Once again, thanks everyone for their participation. I think our method this season was very successful, and I hope that success continues. And I encourage more votes, more stock watch proposals, and more discussion as we move forward, as those things can only help the league.
If there's anything you want to know, or request, or change about the stock watch, feel free to reach out to me.
This season, there were 152 players put up for stock watch, 151 of which had their rating changed (Sorry Nurkic; but thanks for inspiring us to leave polls open longer)
The following GMs voted the most:
Rank | GM | Votes | Notes |
1. | Allan Houston | 152 | |
2. | Billy King | 150 | Not confident on JaMychal Green or Andre Roberson? |
3. | John Stockton | 131 | |
4. | Reggie Miller | 130 | So close to sneaking into the top 3 |
5. | Justin Timberlake | 126 | |
6. | Steve Jobs | 102 | |
7. | Paul Pierce | 94 | |
8. | Logan Hough | 80 | |
9. | Brian Scalabrine | 65 | |
10. | Colin Loftin | 61 | |
11. | DJ Jazzy Jeff | 59 | |
12. | Pistol Pete | 49 | |
13. | Brown Cobb | 46 | It's pretty likely some of his votes were lost. |
T14. | Stan Van Gundy | 28 | |
T14. | Gheorghe Muresan | 28 | RIP. Most votes by a non-active member |
T16. | Luke Walton | 25 | |
T16. | Theodore Duncan | 25 | |
18. | Milt Newton | 20 | |
19. | Ricky Spanish | 18 | RIP |
20. | Hanamichi Sakuragi | 17 | RIP |
21. | Kareem Abdul-Jabbar | 13 | |
22. | Keith Van Horn | 8 | |
T23. | Ari Gold | 7 | |
T23. | Ghost Billy King | 7 | I think this might have been Brown Cobb |
25. | Steven Rogers | 6 | |
26. | Doc Rivers | 3 | |
T27. | Tony Stark | 2 | |
T27. | Steven Armstead | 2 | RIP |
29. | Allen Iverson | 1 |
Thanks everyone for their votes. Those of you who neglected to vote, you're encouraged to do so next season! More votes = more accurate results.
The average change was +1.9. It's a little worrying that there's rating inflation, but I think one mitigating factor is that there are more likely to be rating increases than decreases. There were 87 rating increases this season, compared to only 62 decreases. Nurkic, again, didn't get enough votes to change. Brandon Bass and Al Horford were voted to remain at the same rating.
Why are there more likely to be increases than decreases? For one, it's always beneficial to stock watch your own guys who are due increases. For decreases, GMs need to target other teams, and that takes a bit more effort, and comes with fear of reprisal. Furthermore, underrated players who are unrostered are picked up and stock watched. Overrated players who are unrostered remain unrostered, with no one bothering to stock watch them.
Also, it's telling that the median change was +1. Since the mean is larger, that means that increases were larger than decreases. Part of that is because there were some <45 players who were increases to rotational level guys, whereas decreases would never go below 45.
Which players had the largest changes in a single stock watch?
Top 5 Biggest Increases | ||||
Rank | Player | Old Rating | New Rating | Change |
1. | Dewayne Dedmon | 34 | 57 | +23 |
T2. | Zaza Pachulia | 40 | 60 | +20 |
T2. | Cole Aldrich | 38 | 58 | +20 |
4. | Norman Powell | 40 | 58 | +18 |
5. | Rodney Hood | 46 | 62 | +16 |
Top 5 Biggest Decreases | ||||
Rank | Player | Old Rating | New Rating | Change |
1. | Carlos Boozer | 64 | 50 | -14 |
T2. | JaVale McGee | 63 | 52 | -11 |
T2. | Jermaine O'neal | 59 | 48 | -11 |
T2. | Marshon Brooks | 60 | 49 | -11 |
T2. | Chris Kaman | 63 | 52 | -11 |
Which GMs deviated from the final result the most? Who were the most negative and who the most positive? Using mean signed deviation, the most generous, by far, voter was Doc Rivers. Granted, he only voted 3 times, so hasn't gotten a feel for it yet. That said, he overrated players, relative to the final votes, by and average of 8 1/3 points.
Following Doc is Tony Stark, who in his 2 votes overrated players by an average of 4.5 points. Steven Rogers was +3 average above the result in his 6 votes. Next is Kareem, averaging +2.6 in his 13 votes. Then we have Hanamichi, RIP, +2.05 in 17 votes. And finally, Steve Rogers, +2 in 6 votes.
The differentials on the other side, of those who were much less generous in their votes, are a bit less drastic. That's mostly because they tended to vote more often. It seems like new guys come in and overrate players, whereas the vets are more cynical. As we can see, the most votes among the top five most generous voters were Hanamichi's 17. You'll notice a difference with this next group.
The most negative voter was Milt Newton, who voted on average 1.6 points less than the result, having voted 20 times. Ignoring Ghost Billy King, bc I don't know who those 7 votes really belong to, I come in next at -1.5. To be honest, I thought I'd top this list, often noticing that I voted a player the lowest. But I voted every time for a total of 152 votes, so outliers don't affect much. In third is Theodore Duncan, who in 25 votes averaged -1.16. Next is Keith Van Horn, -1 in 8 votes. Finally, rounding out the top (or should I say bottom) five is Logan Hough, -.66 in 80 votes.
Once again, thanks everyone for their participation. I think our method this season was very successful, and I hope that success continues. And I encourage more votes, more stock watch proposals, and more discussion as we move forward, as those things can only help the league.
If there's anything you want to know, or request, or change about the stock watch, feel free to reach out to me.