RealGM Greatest Peaks Project---List, Interest, Metathinking thread

[Quotatious]

Seems like really 2 more guys are in contention with Durant and Curry in 2014 and 2015.

Assuming we go for top 50, 2015 Harden will probably make it, too.

[JordansBulls]

Here was the old one.

viewtopic.php?f=64&t=1197732

1. Michael Jordan '91, age 27
2. Shaquille O'Neal '00, age 27
3. Bill Russell '65, age 30
4. Wilt Chamberlain '67, age 30
5. Hakeem Olajuwon '94, age 31
6. Kareem Abdul-Jabbar '77, age 29
7. Larry Bird '86, age 29
8. Magic Johnson '87, age 27
9. Tim Duncan '03, age 26
10. LeBron James '09, age 24
11. Kevin Garnett '04, age 27
12. Bill Walton '77, age 24
13. Julius Erving '76, age 25
14. Oscar Robertson '63, age 24
15. Kobe Bryant '08, age 29
16. David Robinson '95, age 29
17. Dwyane Wade '09, age 27
18. Dirk Nowitzki '11, age 32
19. Patrick Ewing '90, age 27
20. Steve Nash '05, age 30
21. Jerry West '66, age 27
22. Chris Paul '08, age 22
23. Charles Barkley '93, age 29
24. Moses Malone '83, age 27
25. Tracy McGrady '03, age 23
26. Karl Malone '98, age 34
27. Scottie Pippen '95, age 29
28. Penny Hardaway '96, age 24
29. Elgin Baylor '61, age 26
30. Dwight Howard '11, age 25
31. Rick Barry '75, age 30
32. Kevin Durant '12, age 23
33. Kevin McHale '87, age 29


We stopped at 33 because participation was not the same. Realistically I think going maybe to top 15 is the way to go if you want the most participation.

[Mutnt]

Rating some of the elder guys might prove more abstract than what we want for healthy discussions. Take someone like Russell for example. How can we accurately compare Russell's peak when we have almost zero information + his case is made on the defensive end which is even harder to gauge.

[Quotatious]

Rating some of the elder guys might prove more abstract than what we want for healthy discussions. Take someone like Russell for example. How can we accurately compare Russell's peak when we have almost zero information + his case is made on the defensive end which is even harder to gauge.

At any rate, I think Russell was voted way too high. What made him great was his unbelievable consistency on a year-by-year basis, but there's no way he peaked higher than Wilt, Kareem, Hakeem or Duncan.

[Mutnt]

That's what I figured, but the problem is, you can't prove it, because data for Russell is non-existent to vague. It's basically a whole bunch of assuming how impactful Russell's D was at his peak, and it has to be pretty high, because his offense is not impressive.

[PaulieWal]

I am in. Seems like a good way to have some NBA fun until the season actually starts.

[theonlyclutch]

Count me in, seems interesting

[Dr Positivity]

Seems like really 2 more guys are in contention with Durant and Curry in 2014 and 2015.

Assuming we go for top 50, 2015 Harden will probably make it, too.

I think Davis and Westbrook are in play as well

[The-Power]

I'll try to contribute as much as possible to the project but I don't necessarily need a vote since I don't know when I'll have the time to actively participate (might be less than I hope, unfortunately).

[70sFan]

I have important question. Do we count every season from every league (BAA, NBA, ABA)? If so, George Mikan for example is the most dominant player ever (more than Shaq and others) based on peak. But he played in worse league with different rules. Also guys like Gilmore, Hawkins and even Dr J had peaks in ABA, where rules also were different (3 point line) than in NBA. Do this years counts in project?

[Dr Positivity]

I have important question. Do we count every season from every league (BAA, NBA, ABA)? If so, George Mikan for example is the most dominant player ever (more than Shaq and others) based on peak. But he played in worse league with different rules. Also guys like Gilmore, Hawkins and even Dr J had peaks in ABA, where rules also were different (3 point line) than in NBA. Do this years counts in project?

Rating players like Mikan, Cousy, Pettit has historically been divisive and up to the voter. Sometimes we have only included post shot clock players as eligible. My personal take is to rate them based on how good they would be now but if giving the training to adapt to the modern game. I won't be voting for Mikan, Cousy or Pettit in my top 50 even if they could all be starters in modern day.

[thizznation]

Woah, Russell at 3? How did that happen?

Anyways, some great ideas have already been kicked around in this thread and it does seem to be about that time to revisit the Peaks Project once again for the reasons already mentioned.

Sign me up!

[Quotatious]

I have important question. Do we count every season from every league (BAA, NBA, ABA)? If so, George Mikan for example is the most dominant player ever (more than Shaq and others) based on peak. But he played in worse league with different rules. Also guys like Gilmore, Hawkins and even Dr J had peaks in ABA, where rules also were different (3 point line) than in NBA. Do this years counts in project?

Honestly, I'd rather start with the 1954-55 season. Mikan was voted at 24 in the top 100 career list, but we really had a lot of problems with him. He could even be top 10 if we only focused at how good each player was compared to his own era, but it's intellectually dishonest to believe that the early 50s were as strong as the following eras.

ABA is okay, I think (IMO '76 Doc is a borderline top 10 peak ever).

[sp6r=underrated]

I don't think rebounds per possession are useful from that era. There were so many more misses it's kinda pointless to compare. You can do TRB% estimates.

By the same token pts or ast per 100 poss are "not very useful" (as more missed shots means fewer of both per possession). I still think per 100 poss estimates are helpful to put some of those old era players stats into perspective; one just has to be conscious of the fact that reb (even per 100 poss) were higher then, and pts and ast were lower.

Not that your method isn't helpful, too.

Rule e) helps guys with multiple peaks a lot.

I don't think it "helps" guys with multiple peaks, but rather just is not penalizing them for having broad or multiple peaks.

We may need to vote on seasons first. Like, say, we agree on 12 nominees for the opening round and then we have a thread for voting on specific seasons for those players. Then every round we nominate a player and vote for a season. The player with the most nominations gets in the voting pool, and the season chosen most often is the designated one. But when you vote a player in, we already have a season chosen beforehand.

For example, in the first round someone could vote for Shaq (2000) and nominate Dirk (2011) while someone else nominates Dirk (2007).

(That still helps players with multiple peaks, but you still have to be voted in so it's not a glaring issue.)

Again, I see this different (I get the feeling at least Quotatious agrees with me). For instance, if for Dirk we by consensus (beforehand) decide his peak is '07 and tell people you MUST consider '07 Dirk only......we might have a handful of people thinking "well, I'd normally give this spot to Dirk IF I could use '11 version, but since I can't, I'm going to vote for someone else". So thus the method you propose still penalizes guys for having multiple or broad peaks......just not quite as much.

This is problematic because (imo) the WHO is much more important than the WHEN.

For instance when we're discussing and comparing players posters will often make statements such as "I think so and so is a top 10 peak of a all-time" or similar, with no mention of what year they're referring to......because ultimately within the context of the MOST of these discussions regarding career value and legacy it is how high a player peaked that is relevant, not what year said peak occurred.
If the majority agrees that Lebron is a top 5 (or perhaps top 3-4) peak all-time, I don't think he should be relegated to #6 or 7 because no one can agree whether it's '09 or '13 or '12 (ditto for guys like Dirk or Kobe or Wade or Wilt). To me, that sort of misses the point of what a peaks project should be about.

I kind of like that idea we've already hashed out (determining year AFTER the player earns his spot), unless a bunch of people voice disagreement.

I'm sorry but that's silly. Simply rearranging words and bolding things doesn't make something true.

It's all about voting for how good a player's particular peak was. If someone doesn't value a particular season highly, then that's useful information and a real opinion. But mixing that up and allowing people to vote for multiple seasons, essentially, means you wash away that signal and a multiple peak guy gets a real advantage. And in most cases it'll be like this: a person prefers 2007 Dirk but because 2011 Dirk is on the board the vote will probably just go to 2011 Dirk anyway.

It is in no way penalizing players with multiple peaks. That doesn't make sense. It's not like votes are being split. You can still vote for the player.

I am not going to participate but I want to state that I strongly agree with ascreamaccrosthecourt. The purpose of an all-peak thread is to evaluate individual seasons. Allowing votes for different seasons to be combined is completely inconsistent with the goal of the project.

[RSCD3_]

I think people would still be nominating single years

[Owly]

By the same token pts or ast per 100 poss are "not very useful" (as more missed shots means fewer of both per possession). I still think per 100 poss estimates are helpful to put some of those old era players stats into perspective; one just has to be conscious of the fact that reb (even per 100 poss) were higher then, and pts and ast were lower.

Not that your method isn't helpful, too.




I don't think it "helps" guys with multiple peaks, but rather just is not penalizing them for having broad or multiple peaks.



Again, I see this different (I get the feeling at least Quotatious agrees with me). For instance, if for Dirk we by consensus (beforehand) decide his peak is '07 and tell people you MUST consider '07 Dirk only......we might have a handful of people thinking "well, I'd normally give this spot to Dirk IF I could use '11 version, but since I can't, I'm going to vote for someone else". So thus the method you propose still penalizes guys for having multiple or broad peaks......just not quite as much.

This is problematic because (imo) the WHO is much more important than the WHEN.

For instance when we're discussing and comparing players posters will often make statements such as "I think so and so is a top 10 peak of a all-time" or similar, with no mention of what year they're referring to......because ultimately within the context of the MOST of these discussions regarding career value and legacy it is how high a player peaked that is relevant, not what year said peak occurred.
If the majority agrees that Lebron is a top 5 (or perhaps top 3-4) peak all-time, I don't think he should be relegated to #6 or 7 because no one can agree whether it's '09 or '13 or '12 (ditto for guys like Dirk or Kobe or Wade or Wilt). To me, that sort of misses the point of what a peaks project should be about.

I kind of like that idea we've already hashed out (determining year AFTER the player earns his spot), unless a bunch of people voice disagreement.

I'm sorry but that's silly. Simply rearranging words and bolding things doesn't make something true.

It's all about voting for how good a player's particular peak was. If someone doesn't value a particular season highly, then that's useful information and a real opinion. But mixing that up and allowing people to vote for multiple seasons, essentially, means you wash away that signal and a multiple peak guy gets a real advantage. And in most cases it'll be like this: a person prefers 2007 Dirk but because 2011 Dirk is on the board the vote will probably just go to 2011 Dirk anyway.

It is in no way penalizing players with multiple peaks. That doesn't make sense. It's not like votes are being split. You can still vote for the player.

I am not going to participate but I want to state that I strongly agree with ascreamaccrosthecourt. The purpose of an all-peak thread is to evaluate individual seasons. Allowing votes for different seasons to be combined is completely inconsistent with the goal of the project.

Given "the project", at least this year is being started by trex, it's odd to tell him that their own interpretation is "inconsistent with the goal of" their own project.

It might be to rate individual seasons (though if primarily this, why not allow multiple years per player), or it might be to rate which players were the best (when they were at their respective bests). Each will have their merits and problems. For instance whilst a season is probably a large enough body of work (to make a judgement on), how you played within a context (teammates, coach, injuries, perhaps era etc) doesn't necessarily define you as a player, so people will probably use context from other years anyhow. I think the above rationale justifies it but it could be the case that (some) people would be blinded by accolades and reputation and ignore highly productive years because their names don't "fit" (for instance how deep would you have to go before seeing Terrell Brandon: 25.24 PER, 0.2365 WS/48; Elton Brand 26.55, 0.2294 or John Drew 25.31; 0.2157 or if older eras were allowed - Larry Foust 23.77, 0.2823; Harry Gallatin 24.142, 0.23241 or Kenny Sears 22.58, 0.2562 - not to say these measures are perfect).

This gist of which is ...
1) For good and bad reasons (and with good and bad consequences) it's very difficult to isolate the idea of a specific season from (perception of) the broader career.

2) There are different possible end goals and methodologies for a player peaks project each of which have their respective strengths and weaknesses, merits and problems (the main problem being discussed here being vote splitting between different years of one player, which in first past the post-single vote system can mean a player who splits fans on their peak, even if backers are united that either year is preferable to a third option, doesn't win - if conversely backers think a third year by another person is superior to their non-preferred year it causes problems for a system which selects only based on the player - different voting systems could help with these problem but cause other problems with regard to organisation and focused discussion and debate).

3) In light of point one, of the two end goals (a ranking of seasons, or a ranking of players) the second might be more feasible (people's inclination to use, and be informed by, other years is less harmful to that). And regarding point two, I think the former problem - splitting votes would be more of an issue than the second - what if people really did split votes because they think the other year is substantially worse, and a different player would be preferable to that particular year.

[The-Power]

I am not going to participate but I want to state that I strongly agree with ascreamaccrosthecourt. The purpose of an all-peak thread is to evaluate individual seasons. Allowing votes for different seasons to be combined is completely inconsistent with the goal of the project.

Your approach, however, would make it necessary to consider every individual season which means we would see multiple seasons of Shaq, Jordan or LeBron - for instance - at the top. As I understood it, that's not the way it is supposed to be and it's not the way previous peak-projects were run either. If you want to have every player on the list only once but still force votes to be split between different peak-seasons, players with an uncertain peak would be punished and ranked lower although their peak-production was generally rated much higher. And that would not only be unfair but kind of pointless as well, in my opinion.

Maybe one could add every vote for a player to the respective player's result which determines his ranking. And then one note how the votes were split between different years. So let's assume LeBron gets 12 votes as the third highest peak. Seven voters take '09 James, four take '13 James and one takes '12 James. Every other player gets less votes, but '67 Wilt gets 8 votes (8 overall). In this case, it's unfair to rank Wilt higher since he as a player got less votes for his peak overall and it is very possible, for example, that '09 James would win a poll against '67 Wilt among the same voters because some of the voters who took '13 James as his peak still prefer his '09 version to Wilt. So why not counting all votes for James, saying he had the third highest peak according to voters (since the most people voted James for the third spot, just different seasons) and then put in brackets how the votes were split between different years.

[trex_8063]

I have important question. Do we count every season from every league (BAA, NBA, ABA)? If so, George Mikan for example is the most dominant player ever (more than Shaq and others) based on peak. But he played in worse league with different rules. Also guys like Gilmore, Hawkins and even Dr J had peaks in ABA, where rules also were different (3 point line) than in NBA. Do this years counts in project?

Rating players like Mikan, Cousy, Pettit has historically been divisive and up to the voter. Sometimes we have only included post shot clock players as eligible. My personal take is to rate them based on how good they would be now but if giving the training to adapt to the modern game. I won't be voting for Mikan, Cousy or Pettit in my top 50 even if they could all be starters in modern day.

The underlined portion denotes a touch of recency bias, imo. It implies most (or perhaps all??) old players will be at least a touch worse in the modern game; and perhaps conversely: that most (or all?) modern players would be at least a touch better in the old-era game. Personally, I don't think either of things is necessarily true.

Manu Ginobili, for example, I suspect has his value go down if he plays in the pre-merger league: no 3pt line (takes away one of his most dangerous offensive weapons, as well as congests the paint===>reducing his ability to effective work in traffic); ***extreme officiating limitations on ball-handling (carries, etc====>bye-bye to behind the back dribble, Euro-step, etc)***; hand-checking is allowed; little in the way of conscientious limitation of minutes, decent medical care, or decent footwear (i.e. imagine Manu's injury/durability woes multiplied by an order of 2-4x)........you get the idea.

***those ball-handling restrictions---along with the allowing of hand-checking and lack of 3pt line in many instances---would mitigate a lot of what allows many modern perimeter players to be so effective. I'm talking about guys like Chris Paul, Steve Nash, Steph Curry, Lebron, Wade, etc.


Moving forward in time, there's an awful lot of variation. For examples, I think it's more than apparent that someone like George Mikan would be no where near as dominant in most other eras other than the one he played in (though he's not near as much a "stiff" as some people make him out to be). otoh, players like Wilt Chamberlain or Sam Jones translate forward with very little effect on their level of dominance (imo).


Ultimately, if you're going to use some manner of era portability/translation as part of your peak assessment, I'd suggest considering how they translate across ALL eras.

But for me personally, era portability speculation is merely a very small portion of my criteria. Not that I'm suggesting we take George Mikan's dominance at face-value. One does, at the least, have to consider the strength of the era said dominance took place in. How to rank the various eras is a tricky and divisive topic, though.

[trex_8063]

It's all about voting for how good a player's particular peak was.

I agree. How good a player's peak was is exactly the point. "How good was his peak?" is ultimately just a paraphrasing of "where does his peak rank all-time?" Is Dirk a top 20 peak all-time? Is Lebron a top 5 peak all-time? Is Kareem a top 5 peak all-time? Is Hakeem a top 5 peak all-time? Is Kevin Durant at top 15 peak all-time? etc etc. These are the questions worth answering, imo, and are the questions I (and apparently most of the other participants) am/are interested in the consensus opinion of. "Who belongs where?": that's the question we are interested in answering. And two persons simply do not have to agree on the year for players like Kobe, Lebron, Dirk, Kareem, Karl Malone, etc, to agree on the answer that that question......Which makes the year seem of side-bar importance.

If someone doesn't value a particular season highly, then that's useful information and a real opinion. But mixing that up and allowing people to vote for multiple seasons, essentially, means you wash away that signal and a multiple peak guy gets a real advantage. And in most cases it'll be like this: a person prefers 2007 Dirk but because 2011 Dirk is on the board the vote will probably just go to 2011 Dirk anyway.

Well, that would kinda undermine the importance of agreeing upon a year anyway then, no?

[trex_8063]

I have important question. Do we count every season from every league (BAA, NBA, ABA)? If so, George Mikan for example is the most dominant player ever (more than Shaq and others) based on peak. But he played in worse league with different rules. Also guys like Gilmore, Hawkins and even Dr J had peaks in ABA, where rules also were different (3 point line) than in NBA. Do this years counts in project?

Good question. Let's come to an agreement on this right now.

I personally would like to include all of the BAA, NBA, and ABA (which would thus put George Mikan into the mix). If others think Mikan "muddies the waters" too much, please speak up now. I otherwise think we should include everything post-shotclock (including the ABA). Thoughts?