RealGM 50 Highest Peaks Project Thread

[JordansBulls]

Wade also played in 4 games in 2007 but that doesn't mean he was healthy. And 1991 Pistons were almost as much "defending champions" as 2007 Heat.

Anyway, the points are:
- 1994 Hakeem dominated on BOTH ends of the floor they way Jordan 1991 didn't. And MVP+DPOTY+FINALS MVP is one of the evidence
- 1994 Rockets faced tougher competition in the playoffs than 1991 Bulls
- 1994 Hakeem had less support than 1991 MJ
- 1994 Hakeem dominated one of the best centers of all time in H2H matchup; 1991 MJ didn't do so to someone similar at his position.
- overall competition in the paint was tougher then than on the perimeter, so Hakeem's productions is more impressive

Pistons made the Conference Finals and upset the #2 seed in the process. Miami lost in round 1 in 2007 by a sweep against a team that had no allstar. Wade got outplayed by Luol Deng in round 1 vs the Bulls which is why the Heat lost.

1. Check (I can agree)
2. Disagree (Bulls beat the 2x defending champions and also the Lakers facing the 2nd best player in the league in Magic Johnson the guy who won 2 MVP's in a row prior to 1991)
3. About the same, both players were the only allstars on there team.
4. MJ went against Magic. Hakeem went against Ewing.
5. Hakeem was 2nd in PER in the playoffs, was 5th in WS/PER 48 minutes, was 1st in win shares. Hakeem was 3rd in Win Shares in the regular season, wasn't even top 5 in WS/PER 48 minutes, and was 3rd in PER (5 points behind Robinson).
MJ was #1 in each of PER , WS and WS/PER minutes 48 minutes in the season and playoffs by a significant margin and literally just like Hakeem was the only allstar on his team that season.

[JordansBulls]

Regarding MJ 1991 vs Shaq 2000


MJ in the season = 20.30 Win Shares. Shaq = 18.65
MJ in the season = 31.63 PER, Shaq = 30.65


MJ in the playoffs = 4.77 Win Share, Shaq = 4.67
MJ in the playoffs = 32.04 PER, Shaq = 30.45


MJ in the finals = 31.5 PER and Shaq = 31.1


http://sports.espn.go.com/nba/playoffs2008/columns/story?columnist=hollinger_john&page=FinalsPerformances-7
http://sports.espn.go.com/nba/playoffs2008/columns/story?columnist=hollinger_john&page=FinalsPerformances-8

So these two were neck and neck.

MJ had over a 31.5 PER in the season, playoffs and finals in 1991

[ardee]

Regarding MJ 1991 vs Shaq 2000


MJ in the season = 20.30 Win Shares. Shaq = 18.65
MJ in the season = 31.63 PER, Shaq = 30.65


MJ in the playoffs = 4.77 Win Share, Shaq = 4.67
MJ in the playoffs = 32.04 PER, Shaq = 30.45


MJ in the finals = 31.5 PER and Shaq = 31.1


http://sports.espn.go.com/nba/playoffs2008/columns/story?columnist=hollinger_john&page=FinalsPerformances-7
http://sports.espn.go.com/nba/playoffs2008/columns/story?columnist=hollinger_john&page=FinalsPerformances-8

So these two were neck and neck.

MJ had over a 31.5 PER in the season, playoffs and finals in 1991

You can't judge this based ONLY on PER and WS

I agree MJ 1991 was better than Shaq 2000 (the guy was not on his usual level in the toughest series of the year, against the Blazers, and actually pretty poor at times in games 6 and 7, the two biggest games of the year for him), but seriously, there are many MANY better arguments than convoluted stats like those...

[JordansBulls]

Regarding MJ 1991 vs Shaq 2000


MJ in the season = 20.30 Win Shares. Shaq = 18.65
MJ in the season = 31.63 PER, Shaq = 30.65


MJ in the playoffs = 4.77 Win Share, Shaq = 4.67
MJ in the playoffs = 32.04 PER, Shaq = 30.45


MJ in the finals = 31.5 PER and Shaq = 31.1


http://sports.espn.go.com/nba/playoffs2008/columns/story?columnist=hollinger_john&page=FinalsPerformances-7
http://sports.espn.go.com/nba/playoffs2008/columns/story?columnist=hollinger_john&page=FinalsPerformances-8

So these two were neck and neck.

MJ had over a 31.5 PER in the season, playoffs and finals in 1991

You can't judge this based ONLY on PER and WS

I agree MJ 1991 was better than Shaq 2000 (the guy was not on his usual level in the toughest series of the year, against the Blazers, and actually pretty poor at times in games 6 and 7, the two biggest games of the year for him), but seriously, there are many MANY better arguments than convoluted stats like those...

I was just throwing the numbers together to show the overall production of the two.

I think I have Wilt's 1967 5th but that is mainly due to it being like his 6th best season PER wise. Also in the playoffs he was the 2nd leading scorer on his team and in the finals he was the 5th leading scorer on the team.

Remember scoring was probably Wilt's greatest strength and to be 5th in it on the team in the finals is major.

http://webuns.chez-alice.fr/finals/1967.htm#

PHI. G FG-FGA FT-FTA REB AST PF PTS
Greer 6 59-148 38-46 48 37 23 156
Walker 6 46-102 48-62 53 20 23 140
Jones 6 50-110 21-28 21 32 22 121
Cunningham 6 48-107 22-40 34 18 27 118
Chamberlain 6 42-75 22-72 171 41 16 106

[Doctor MJ]

While I'm not knowledgeable in basketball nor do I intend on contributing on the actual project, let me ask about a certain scenario.


Let's say there are 4 people voted on for a certain position. 20 votes total.
A gets 8
B gets 7
C gets 3
D gets 2.
A obviously won, but...
Let's say people who voted for C/D didn't agree that A was better than B. In fact, if there was a revote with only A and B the results would be
A gets 9
B gets 11.

So I suppose the overall scenario is, if more people think one peak is better than the rest, but on a whole more people believe one peak is better than the other, how do you judge that?

realbig gave a good response about how practically this won't be a huge issue.

Theoretically what you're getting into is some of the intricacies of Voting Theory where in the end, there is no fool proof solution. The strongest solution we could probably use would be Instant Runoff Voting where each person would give not a single vote but a ranked preference list. However, to do that well would require significant overhead that no one running a project has ever been willing to do here (others have volunteered to do the tallying, but I've never been satisfied with their proposals).

In the end, the most important thing to remember is that the best part of these projects is the discussion. These are not definitive lists, but rather engines of enjoyment and learning. All that said, I do think the results are pretty indicative of beliefs for a variety of reasons relating to practical implementation.

[Dr Positivity]

I see flaws with the system of straight up voting as well. I think putting into plays scenarios like "2000 Shaq - 15 votes, 94 Hakeem - 12 votes, 95 Hakeem - 5 votes, winner: Shaq" as outlined in the last thread, where Shaq wins but Hakeem's peak likely wins if all the 94/95 voters had to pick one, is definitely undesirable.

What I would go for is a pre vote to see which year is the "peak" year for every player, settling 94 vsv 95 Hakeem, 91 vs 93 Jordan, 71 vs 77 Kareem, 12 vs 09 Lebron, etc. Then when we get to the real debates the only options are eg. 94 Hakeem, 91 Jordan, 71 Kareem, 09 Lebron of those 4 players

[Dr Positivity]

Or another option if we don't want to do to the pre-vote, a system similar to IRV but not directly - make it so everyone gives a top 3-5, but the run-off only applies if it's the same player who's #1

For example in the '2000 Shaq - 15 votes, 94 Hakeem - 12 votes, 95 Hakeem - 5 votes' scenario I listed

Voter A's Ballot
1. 1995 Hakeem
2. 1994 Hakeem
3. 2000 Shaq

In this scenario when 95 Hakeem is "eliminated", his vote transfers to 1994 Hakeem because it's the same player and he clearly thinks Hakeem's peak is higher than Shaq's

But in the case of

Voter B's Ballot
1. 1986 Larry Bird
2. 1994 Hakeem
3. 2000 Shaquille O'Neal

His vote counts for Bird, it doesn't run-off to 1994 Hakeem and isn't used to decide Shaq vs Hakeem

I think the most simple way to do it though is a pre-vote where every player is assigned 1 year. It could just be in one thread that runs for a week. Yes there wouldn't be as much discussion/opportunity for perfect accuracy as in individual threads, but does it really change the project much whether 91 or 93 Jordan is his representative or 09 or 12 Lebron? I think the important part is going to be when separate players are pitted against each other, not which individual year represents their peak

[C-izMe]

I like that idea. It will end up with better results than trying to get everyone to notice who's ahead to change their votes.

[Vinsanity420]

Agree with Dr. Mufasa's system. There should be more of an emphasis on debating player vs player peak, rather than spending the bulk of your time nitpicking specific years.

[ThaRegul8r]

Ok guys, tell me why Hakeem 1994 shouldn't be no 1? (And I hope answer would be different than PER or WS...)

Probably because some people (including me) are confused as to whether 1994 or 1995 was his peak year impact-wise.

'94 was the MVP season, yet '95 is the legendary playoffs which had everyone reassess Hakeem's place in history.

[ardee]

Ok guys, tell me why Hakeem 1994 shouldn't be no 1? (And I hope answer would be different than PER or WS...)

Probably because some people (including me) are confused as to whether 1994 or 1995 was his peak year impact-wise.

'94 was the MVP season, yet '95 is the legendary playoffs which had everyone reassess Hakeem's place in history.

Indeed... I'm not sure people appreciate the degree of difficulty Hakeem was up against in that postseason.

To whoever said he had more help, that's false, his team was significantly worse than the previous year. Drexler was a shadow of his former self, Hakeem had to carry the offensive load pretty much by himself, and was also defending a frontcourt with an elite player in practically every series (Malone, Barkley, Robinson and Shaq).

It was like LeBron this year, only if he was the only All Star on his team, and had to go through a slightly past his prime Kobe, a prime Wade, a peak Doc, and a prime Bird all in the same postseason.

[lorak]

Not true ardee - Drexler played on high level in 1995 playoffs, offensively he might even be better than Hakeem that post season.

Also, Rockets role players were really in the zone, hitting clutch shot after clutch shot. Part of it was of course because Hakeem and Drexler draw defensive attention, but also many of these shots were simply amazing, through the hands, last second and so on. That run was magical for all Rockets players.

[An Unbiased Fan]

I won't have time to do this project, but it does look promising. My only question is why not choose a 3 year "Peak" as opposed to 1 year. For Hakeem that would be 93-95', MJ would be 91-93, Bird would be 84-86, Kobe would be 06-08', Shaq would be 00-02', and so on...

The few times peaks have been ranked on the PC board, this is typically the method used. It gives a bigger sample size, and is a bit more reflective of a player.

For example, in the NFL, Jamal Lewis rushed for 2066 yds, at 129.1 yds per game, 5.3 per carry in 03'. That's #2 all-time, and if we went by peak year, then many could make a case that he had the 2nd greatest "peak" behind Dickerson. However, if we look at his 3 year peak, we would see that 2003 was an outlying season, and that he was never consistently at that level.

If this project is ranking the highest "peaks", as opposed to the best individual seasons, then perhaps a 3-year sample would work better.

[colts18]

Hakeem's teammates had over a .600 TS% vs. the Magic in the finals. They were doing their jobs

[Kobe 62 Mavs 61]

I won't have time to do this project, but it does look promising. My only question is why not choose a 3 year "Peak" as opposed to 1 year. For Hakeem that would be 93-95', MJ would be 91-93, Bird would be 84-86, Kobe would be 06-08', Shaq would be 00-02', and so on...

The few times peaks have been ranked on the PC board, this is typically the method used. It gives a bigger sample size, and is a bit more reflective of a player.

For example, in the NFL, Jamal Lewis rushed for 2066 yds, at 129.1 yds per game, 5.3 per carry in 03'. That's #2 all-time, and if we went by peak year, then many could make a case that he had the 2nd greatest "peak" behind Dickerson. However, if we look at his 3 year peak, we would see that 2003 was an outlying season, and that he was never consistently at that level.

If this project is ranking the highest "peaks", as opposed to the best individual seasons, then perhaps a 3-year sample would work better.

I think that might be the difference between "prime" and "peak".

Maybe it's semantics, but I've always thought a peak would be one year and a prime would be their best three, or something like that.

[An Unbiased Fan]

I won't have time to do this project, but it does look promising. My only question is why not choose a 3 year "Peak" as opposed to 1 year. For Hakeem that would be 93-95', MJ would be 91-93, Bird would be 84-86, Kobe would be 06-08', Shaq would be 00-02', and so on...

The few times peaks have been ranked on the PC board, this is typically the method used. It gives a bigger sample size, and is a bit more reflective of a player.

For example, in the NFL, Jamal Lewis rushed for 2066 yds, at 129.1 yds per game, 5.3 per carry in 03'. That's #2 all-time, and if we went by peak year, then many could make a case that he had the 2nd greatest "peak" behind Dickerson. However, if we look at his 3 year peak, we would see that 2003 was an outlying season, and that he was never consistently at that level.

If this project is ranking the highest "peaks", as opposed to the best individual seasons, then perhaps a 3-year sample would work better.

I think that might be the difference between "prime" and "peak".

Maybe it's semantics, but I've always thought a peak would be one year and a prime would be their best three, or something like that.

I would say primes last a bit longer than 3 years(more like 6-9), and speak more to longevity. Peaks show a player at their optimal level.

[Rapcity_11]

Peaks show a player at their optimal level.

Which for some guys is only 1 year.

I understand you want to get rid of the luck factor but extending it to multiple years will become too similar to the top 100 list. Part of the fun is seeing where guys like Walton and T-Mac fit in.

[An Unbiased Fan]

Peaks show a player at their optimal level.

Which for some guys is only 1 year.

I understand you want to get rid of the luck factor but extending it to multiple years will become too similar to the top 100 list. Part of the fun is seeing where guys like Walton and T-Mac fit in.

I guess my main problem is that you have to separate 00' Shaq from 01' Shaq, or 91' MJ from 92' MJ. There are things in both of those years that tell us a lot about their optimal playing level. You could go 01-03' for Tmac, but as you pointed out, a guy like Walton would likely fall short of 3 great years, though 76-78 could be used.

But I do see your point. Perhaps that would be a bit too much like the Top 100.

[Doctor MJ]

I won't have time to do this project, but it does look promising. My only question is why not choose a 3 year "Peak" as opposed to 1 year. For Hakeem that would be 93-95', MJ would be 91-93, Bird would be 84-86, Kobe would be 06-08', Shaq would be 00-02', and so on...

The few times peaks have been ranked on the PC board, this is typically the method used. It gives a bigger sample size, and is a bit more reflective of a player.

For example, in the NFL, Jamal Lewis rushed for 2066 yds, at 129.1 yds per game, 5.3 per carry in 03'. That's #2 all-time, and if we went by peak year, then many could make a case that he had the 2nd greatest "peak" behind Dickerson. However, if we look at his 3 year peak, we would see that 2003 was an outlying season, and that he was never consistently at that level.

If this project is ranking the highest "peaks", as opposed to the best individual seasons, then perhaps a 3-year sample would work better.

Well, I think it's pretty telling that the example you're thinking of is from another sport. I don't think a 1 year peak is the best way to judge every athlete. However I think basketball is a consistent enough sport that it works.

I also think that in the cases where there would be a significant difference between a 1 year period and a 3 year period, I don't see the advantage of ignoring the 1 year peak. Doing so would exclude a guy like Walton obviously, and what would be the point in that? Moreover, doing so would seriously diminish Wilt's candidacy at least in my eyes, leading to the same battles we've had so many times before.

Also in principle, I always like to minimize the arbitrariness of any criteria I use. 1 season is the smallest complete unit of basketball. It's a logical place to evaluate. If we use a 3 season standard, why not 2? and why not 4? or 5?

With all this said, getting back to what you say about Lewis, I'd think the main reason it would be unsatisfactory in someone's eyes to use a 1-year peak judgement on him would be that you thought that there was some form of luck involved in that year that makes him look better than he is. In football, where a player is dependent on so many other players, this is a very reasonable thing to argue, and it's not necessarily irrelevant in basketball either...

However, we're certainly not asking you to ignore luck here. If a voter thinks that a player's true abilities got overrated based on some outside factor in his big season, they should proclaim so loudly and vote their conscience.

[Doctor MJ]

I guess my main problem is that you have to separate 00' Shaq from 01' Shaq, or 91' MJ from 92' MJ.

I get that. I am anticipating that people will be pragmatic about such debates and not take them as seriously as, say, '91 vs '88 Jordan. However, if such debates appear to be causing problems in the voting, or are actually bogging down threads into "angels on a pin" levels of silliness, I'm confident we can make adjustments to clear the pipes.