Greatest Peaks Project (2022): #26 - 2019-20 Anthony Davis

[capfan33]

1. 2020 Davis
I don't buy his 2020 performance as at all being reflective of his actual level as a player. But even previous versions of Davis, if you buy his playoff performance level in a limited sample, would still probably be 1st for me. Simply put, no one left has the level of 2-way versatility that Davis has. I seriously considered Ewing but ultimately feel a bit more comfortable with Davis' versatility over Ewings incredible rim protection.

2. 90 Ewing
I'm not very familiar with Ewing overall, but I do know he was a historically great defensive anchor who anchored some of the greatest defenses ever. AEngima's posts in the previous thread are largely the reason for this vote, as he showed how 1990 for Ewing was actually a legitimate outlier for him due to an improved offensive skillset while still maintaining his athleticism, and I suspect that the better defensive results in 93 and 94 are more due to roster construction and coaching than Ewing himself. In a general sense, I like his skillset quite a bit, a legitimate top-10 defensive anchor ever who also had outside shooting touch and athleticism to play in the pick and roll. It plays well in many teams in any era, and as such I'm unexpectedly voting for Ewing here.

3. 07 Nash
Nash is probably a top-5ish offensive player ever with outlier impact numbers and team results. He was an incredible passer and shooter, top 10 all time in both (at worst). His results seem relatively resilient in the postseason and he did so on 2 pretty different teams. I would say Harden is probably my next pick but I'm not sold at all on that, and I like Nash's offensive significantly more than Hardens.

[falcolombardi]

1. 2020 Davis
I don't buy his 2020 performance as at all being reflective of his actual level as a player. But even previous versions of Davis, if you buy his playoff performance level in a limited sample, would still probably be 1st for me. Simply put, no one left has the level of 2-way versatility that Davis has. I seriously considered Ewing but ultimately feel a bit more comfortable with Davis' versatility over Ewings incredible rim protection.

2. 90 Ewing
I'm not very familiar with Ewing overall, but I do know he was a historically great defensive anchor who anchored some of the greatest defenses ever. AEngima's posts in the previous thread are largely the reason for this vote, as he showed how 1990 for Ewing was actually a legitimate outlier for him due to an improved offensive skillset while still maintaining his athleticism, and I suspect that the better defensive results in 93 and 94 are more due to roster construction and coaching than Ewing himself. In a general sense, I like his skillset quite a bit, a legitimate top-10 defensive anchor ever who also had outside shooting touch and athleticism to play in the pick and roll. It plays well in many teams in any era, and as such I'm unexpectedly voting for Ewing here.

3. 07 Nash
Nash is probably a top-5ish offensive player ever with outlier impact numbers and team results. He was an incredible passer and shooter, top 10 all time in both (at worst). His results seem relatively resilient in the postseason and he did so on 2 pretty different teams. I would say Harden is probably my next pick but I'm not sold at all on that, and I like Nash's offensive significantly more than Hardens.

I recommend watching this game

https://youtu.be/E_RNkZ_losY

Shows ewing strenghrs and weaknesses pretty well may do a write up on ewing a bit later when i compare him with barkley

[NBA4Lyfe]

nash over cp3 and harden lol

[AEnigma]

shaq over cp3 and harden lol

[NBA4Lyfe]

shaq over cp3 and harden lol

i said it before and i'll say it again. the best way to do an all-time rankings list without bias is too try and convince basketball reference.com to rank their top 50 players.

[AEnigma]

I’m just lazy to make a post and Moses is gonna win this round anyway ngl lol

Yeah in the end I wasn't sure of who to fill out my ballot with(Nash or McGrady probably, the Ewing arguments made me less confident in having Moses there like I did originally) with the time I have, and it looks like Moses is getting in anyways - i'll maybe have a decision on that before the deadline but it's looking like I might as well wait until next round

Curious how either of you feel about Dwight. His offence really is unusual for these types of projects, and I have never been too comfortable in assessing it. We see guys like Gobert have significant schematic impact on their teams’ offence despite falling short of Dwight’s rim pressure, but I am not sure Dwight quite evidenced that same commitment to smart and involved team play on that end. Then again, maybe that is more a consequence of era.

[Proxy]

Curious how either of you feel about Dwight. His offence really is unusual for these types of projects, and I have never been too comfortable in assessing it. We see guys like Gobert have significant schematic impact on their teams’ offence despite falling short of Dwight’s rim pressure, but I am not sure Dwight quite evidenced that same commitment to smart and involved team play on that end. Then again, maybe that is more a consequence of era.

This is how I feel about Dwight mostly, I wouldn't call myself particularly high on his offense as I don't really see the evidence of that either, but really the one thing that gives me pause is his absurd foul drawing. Dwight might legit be t3-5 ish or so all-time in this regard(though hack-a-Dwight plays a part, similar to someone like Shaq) his effect on team FT rates even past his absolute peak just seem so big to the point where it probably ended up being a pretty clear + add on offense.
http://nbashotcharts.com/rapm3?id=550492930

It could be a consequence of era, but i'm not sure how much I really believe in Dwight doing that in other eras either - doesn't really strike me as that type of player but who knows really. Maybe it played into the oddly somewhat high turnover rate a bit for a player doing what he was expected to do(though Rudy also has a pretty bad turnover problem too - he was like bottom in the league in turnovers per touch alongside players like PG and Drummond but I think his biggest problem is maybe his hands)

[Proxy]

Yeah unless he recently updated his CORP calculations I do not think that is right. Because he definitely has Nash higher than Paul on his top 40 page, and then a copy I have of some values taken from Patreon data has them at the same CORP value despite giving 2007 Nash worse “port” and somehow a worse health grade than 2015 Paul (good luck selling people on that lol). On the same note, Ewing was at the same level as Malone and Barkley, not lower.

Yeah he changed those Nash evals very soon after publishing those articles because less than 2 years later he dropped him

His 2010 Nash eval now has him at 16.5% CORP(2007 would be the same without the injury thing), while his 2014 Paul eval is at 17.2%(and 2016 without the injury tax would be even higher as he had that version as a +5.75 player and 2014 Paul as a +5.5).

He originally pegged 2007 Nash as a +6.5 Offense player(-2 port), and had since then dropped him down to +6(again, I think this update came within like a year max) - that's probably why the % looks higher in the backpicks profile. He also changed his Paul evals from originally having him at -1 port to neutral after ths Houston stint(same +/- eval)

[AEnigma]

2010? Huh, that is… unusual.

I have always been more interested in his plus/minus assessments than his relatively arbitrary port penalisations, but I guess I will need to wait for that Backpicks podcast update to hear why he suddenly reassessed Nash’s peak plus/minus value and shifted to 2010 as his peak.

[falcolombardi]

2010? Huh, that is… unusual.

I have always been more interested in his plus/minus assessments than his relatively arbitrary port penalisations, but I guess I will need to wait for that Backpicks podcast update to hear why he suddenly reassessed Nash’s peak plus/minus value and shifted to 2010 as his peak.

I would guess the kinda arbitrary port evaluations you mention may be why he ended valuing 2010 higher

[falcolombardi]

shaq over cp3 and harden lol

i said it before and i'll say it again. the best way to do an all-time rankings list without bias is too try and convince basketball reference.com to rank their top 50 players.

That may be less human bias, but you elininate the human assesment element and even a lot of the statistical elements bballref doesnt capture

It would "unbiased" but more arbitrary

[Proxy]

2010? Huh, that is… unusual.

I have always been more interested in his plus/minus assessments than his relatively arbitrary port penalisations, but I guess I will need to wait for that Backpicks podcast update to hear why he suddenly reassessed Nash’s peak plus/minus value and shifted to 2010 as his peak.

I would guess the kinda arbitrary port evaluations you mention may be why he ended valuing 2010 higher

Nah it's just because of health, they have the same evaluation otherwise for him.

Also I don't think you can really talk about his plus/minus assessments without mentioning the scalability curve - it don't believe it's like he just comes up with one and then decides the "tax" after, he has a whole article on that stuff actually and Nash was an example he gave in a Q&A once about how you can just modify values to come out to the same CORP%(eg you could view Nash as neutral port but the +/- eval would change) - but yeah it's still just arbitrary and an explanation for the Nash change would be good. I still believe his explanations behind his CORP formula inputs are just weird and lacking. https://backpicks.com/2021/06/04/scalability-curves-and-greatest-peaks-corp/

[NBA4Lyfe]

shaq over cp3 and harden lol

i said it before and i'll say it again. the best way to do an all-time rankings list without bias is too try and convince basketball reference.com to rank their top 50 players.

That may be less human bias, but you elininate the human assesment element and even a lot of the statistical elements bballref doesnt capture

It would "unbiased" but more arbitrary

the human element has bias. for instance harden and cp3 are very polarizing players. taking the human element out completely fixes that and we get a fair ranking. Too many times people are just afraid of comparing advanced stats and basic stats instead of which style they prefer or which player they like. Every player has stats, you wanna make comparisons head over to basketball reference and compare. For instance harden peaked higher than both kobe and wade and he is not yet on this list. The human element is too bias

[MyUniBroDavis]

i said it before and i'll say it again. the best way to do an all-time rankings list without bias is too try and convince basketball reference.com to rank their top 50 players.

That may be less human bias, but you elininate the human assesment element and even a lot of the statistical elements bballref doesnt capture

It would "unbiased" but more arbitrary

the human element has bias. for instance harden and cp3 are very polarizing players. taking the human element out completely fixes that and we get a fair ranking. Too many times people are just afraid of comparing advanced stats and basic stats instead of which style they prefer or which player they like. Every player has stats, you wanna make comparisons head over to basketball reference and compare. For instance harden peaked higher than both kobe and wade and he is not yet on this list. The human element is too bias

Using code trying to parse out peaks with only bball ref would be horrible on a different scale lol

[falcolombardi]

2010? Huh, that is… unusual.

I have always been more interested in his plus/minus assessments than his relatively arbitrary port penalisations, but I guess I will need to wait for that Backpicks podcast update to hear why he suddenly reassessed Nash’s peak plus/minus value and shifted to 2010 as his peak.

I would guess the kinda arbitrary port evaluations you mention may be why he ended valuing 2010 higher

Nah it's just because of health, they have the same evaluation otherwise for him.

Also I don't think you can really talk about his plus/minus assessments without mentioning the scalability curve - it don't believe it's like he just comes up with one and then decides the "tax" after, he has a whole article on that stuff actually and Nash was an example he gave in a Q&A once about how you can just modify values to come out to the same CORP%(eg you could view Nash as neutral port but the +/- eval would change) - but yeah it's still just arbitrary and an explanation for the Nash change would be good. I still believe his explanations behind his CORP formula inputs are just weird and lacking. https://backpicks.com/2021/06/04/scalability-curves-and-greatest-peaks-corp/

I dont want you to give me the full thingh but do you have a rough outline of what ben corp formula accounts for? Not the full formula lol, just what kind of thinghs does he consider for his corp or port formulas?

I am trying to get an idea of what does he take into account to get those numbers

I dont want to make a patreon account only for this lol

[Proxy]

I dont want you to give me the full thingh but do you have a rough outline of what ben corp formula accounts for? Not the full formula lol, just what kind of thinghs does he consider for his corp or port formulas?

I am trying to get an idea of what does he take into account to get those numbers

I dont want to make a patreon account only for this lol

I'm just gonna copy paste sections parts from the overall article word for word lol, I pay him for it so w/e. Most of the article is talking about things like scalability curves/odds of redundancies/buildability that I could send if you need them(maybe that's what you meant?).

There’s also the issue of price. The economic probabilities of CORP are a pandora’s box that’s even more complex than modeling playoff opponents or scalability curves. Secondary offensive players — or even complementary ones like Draymond — are almost always cheaper options. Volume scorers tend to demand the most money, therefore it’s unlikely to find a ton of them on the same team. My current CORP calculator estimates the odds of playing on varying teams based on historical data, but it doesn’t account for the salary cap of a given team and the player’s contract demands.

You might be thinking, “quick, go create a calculator with econ probabilities!” But it’s not that simple. There isn’t always a one-to-one relationship between value and cost, so do we reward players when the market overrates or underrates their value? Peak Steph Curry was a max player on a discount contract. Draymond Green too. Heck, Magic Johnson and Michael Jordan were members of the underpaid club themselves.12

I’ve been told that it’s nearly impossible for superstars to play on high-end teams, and that’s why floor-raising is so important. I’d argue that before unrestricted free-agency (1988-89), most title teams were “superteams” that were formed with some sort of luck. Bird was picked a year early, the Lakers won a coin flip for Magic and ended up with the first pick after winning the title in 1982 (!), and the 76ers paired the last two MVPs with two other All-Stars in 1983.13 Even the ’90s Bulls were loaded, with Scottie Pippen finishing third in MVP voting in his one year without MJ. The 2017 Warriors are the most extreme example, but economic flukes aren’t that uncommon.

The 2012 Thunder prematurely broke up three future MVP’s. Economics were obviously a factor…but so was fit! James Harden came off the bench because he was No. 2 in their ball-dominant hierarchy. If Harden or Westbrook played a different role at the same level of talent, OKC might have stayed together longer and won a few championships in the process. So whether a team signs their other stars, trades for them or develops them at home, avoiding redundancy makes a world of difference.

Overall, I don’t think these economic complexities distort the current CORP calculation that much. Most teams win by stacking good players, and dynasties form when those stars fit together to create a large competitive advantage.

----

The last component of championship equity is health. From a results-oriented standpoint, regular season health doesn’t matter that much. Missing 20 or even 40 games can hurt playoff seeding slightly, but there are a few things to remember:

• Home court advantage helps, but being the better team is more important than being the home team for an extra game, unless the teams are relatively close.

• Most good teams have to defeat two other legit contenders, often in the Conference Finals and the Finals, regardless of their seed.

• Really good teams need good supporting casts, so most of these teams should be able to make the playoffs without their star player.

Because of the first two observations, the difference between the first seed or sixth seed doesn’t change a team’s championship odds that much, assuming they’re really good at full strength. (Think ’01 Lakers.) Any team worth its salt is usually going to make the postseason without a key player, unless injuries ravage the roster. So the real question is how to handle missed time in the playoffs, where a few losses end a season.

Right now, I punish players for missing playoff games. The later in the postseason they miss time, the greater punishment, because that’s precisely where the opponents are most challenging. But there are issues with this approach. The playoffs are physically taxing, so players on lottery teams don’t have the chance to injure themselves after the regular season! In essence, fragile players like Chris Paul and Kawhi Leonard are punished by advancing further down the road. This doesn’t strike me as fair when comparing player value.

For most players, this largely evens out over their careers. Some are lucky enough to “time” their injuries perfectly and never miss playoff time (Shaq), while others have a decade of nagging injuries that catch up to them in the postseason (Paul). But in general, guys who miss time in the regular season also miss time in the playoffs — there is a predictive nature to the larger regular season in that sense.

A more probabilistic-based approach to injuries is probably more fair. Maybe I should assume that a player misses some number of playoff games in a given season based on his injury history across adjacent seasons at that point in his career. Depending on style of play and the nature of the injury, that might mean retrospectively dinging someone for health, even though they were quite healthy until that point in their career. (eg Derrick Rose) I’m not sure how to implement this yet, but a probabilistic approach could remain objective while reducing the “can’t get injured in Cancun” problem. Perhaps this will be part of a CORP update in the future.

---

For now, it’s the same results-oriented approach to health. In the table below, you’ll find the latest valuations from Greatest Peaks along with:

• A base +/- estimate with offensive and defensive estimates. Think of this as performance on a slightly-below .500 team, which is the middle part of the bell curve. (Removing players from average teams makes them slightly below average.)

• A scalability curve (portability) for a player’s offense, based on the base estimate. There are five different shapes, with -2 curves representing a steeper decline in value, and +2 holding value more on better and better teams. (None of these 5 curves are as extreme as the Iverson/Draymond examples from above.)

• Health, based on their regular season and playoff games played (GP)

• CORP ranges

The ranges are the newest addition, and represent a low-end evaluation and a high end valuation. The high-end of the range is how I view a player’s offense and defense in the best reasonable light at the same time, so if ramping up offense hurts defense, that’s taken into account. All valuations are based off my CORP calculator, with minor mental curving if I don’t think the five generic scalability curves apply.

Finally, let’s take everything we’ve outlined and apply it to Jordan and LeBron. Excluding health, the CORP calculator is solving for a player’s title equity based on three key inputs:

• Base offensive value.

• The scalability curve associated with that base value.

• Base defensive value. (The calculator views defense as scaling equally well for all players, since defense generally scales well.)

LeBron’s 2009 has a base +/- valuation of +7.38 with a -2 scaling curve (steeper decline) on offense. Jordan’s 1990 is +7.13 with a neutral scaling curve (more gradual decline). In other words, I think LeBron is going to be slightly better helping an average team, but he doesn’t hold that value as well as Jordan does as his team quality improves. So at some point — maybe by the time we reach, say, a +3 team? — Jordan will start to have greater impact.

Remember, it’s impossible to describe generalized impact without a scalability curve. We could achieve a similar CORP value with a different combination of base values and scalability curves, which is why LeBron’s 2013 and Jordan’s 1991 have nearly identical CORP, despite having different base values.

[DraymondGold]

I dont want you to give me the full thingh but do you have a rough outline of what ben corp formula accounts for? Not the full formula lol, just what kind of thinghs does he consider for his corp or port formulas?

I am trying to get an idea of what does he take into account to get those numbers

I dont want to make a patreon account only for this lol

I'm just gonna copy paste sections parts from the overall article word for word lol, I pay him for it so w/e. Most of the article is talking about things like scalability curves/odds of redundancies/buildability that I could send if you need them(maybe that's what you meant?).

There’s also the issue of price. The economic probabilities of CORP are a pandora’s box that’s even more complex than modeling playoff opponents or scalability curves. Secondary offensive players — or even complementary ones like Draymond — are almost always cheaper options. Volume scorers tend to demand the most money, therefore it’s unlikely to find a ton of them on the same team. My current CORP calculator estimates the odds of playing on varying teams based on historical data, but it doesn’t account for the salary cap of a given team and the player’s contract demands.

You might be thinking, “quick, go create a calculator with econ probabilities!” But it’s not that simple. There isn’t always a one-to-one relationship between value and cost, so do we reward players when the market overrates or underrates their value? Peak Steph Curry was a max player on a discount contract. Draymond Green too. Heck, Magic Johnson and Michael Jordan were members of the underpaid club themselves.12

I’ve been told that it’s nearly impossible for superstars to play on high-end teams, and that’s why floor-raising is so important. I’d argue that before unrestricted free-agency (1988-89), most title teams were “superteams” that were formed with some sort of luck. Bird was picked a year early, the Lakers won a coin flip for Magic and ended up with the first pick after winning the title in 1982 (!), and the 76ers paired the last two MVPs with two other All-Stars in 1983.13 Even the ’90s Bulls were loaded, with Scottie Pippen finishing third in MVP voting in his one year without MJ. The 2017 Warriors are the most extreme example, but economic flukes aren’t that uncommon.

The 2012 Thunder prematurely broke up three future MVP’s. Economics were obviously a factor…but so was fit! James Harden came off the bench because he was No. 2 in their ball-dominant hierarchy. If Harden or Westbrook played a different role at the same level of talent, OKC might have stayed together longer and won a few championships in the process. So whether a team signs their other stars, trades for them or develops them at home, avoiding redundancy makes a world of difference.

Overall, I don’t think these economic complexities distort the current CORP calculation that much. Most teams win by stacking good players, and dynasties form when those stars fit together to create a large competitive advantage.

----

The last component of championship equity is health. From a results-oriented standpoint, regular season health doesn’t matter that much. Missing 20 or even 40 games can hurt playoff seeding slightly, but there are a few things to remember:

• Home court advantage helps, but being the better team is more important than being the home team for an extra game, unless the teams are relatively close.

• Most good teams have to defeat two other legit contenders, often in the Conference Finals and the Finals, regardless of their seed.

• Really good teams need good supporting casts, so most of these teams should be able to make the playoffs without their star player.

Because of the first two observations, the difference between the first seed or sixth seed doesn’t change a team’s championship odds that much, assuming they’re really good at full strength. (Think ’01 Lakers.) Any team worth its salt is usually going to make the postseason without a key player, unless injuries ravage the roster. So the real question is how to handle missed time in the playoffs, where a few losses end a season.

Right now, I punish players for missing playoff games. The later in the postseason they miss time, the greater punishment, because that’s precisely where the opponents are most challenging. But there are issues with this approach. The playoffs are physically taxing, so players on lottery teams don’t have the chance to injure themselves after the regular season! In essence, fragile players like Chris Paul and Kawhi Leonard are punished by advancing further down the road. This doesn’t strike me as fair when comparing player value.

For most players, this largely evens out over their careers. Some are lucky enough to “time” their injuries perfectly and never miss playoff time (Shaq), while others have a decade of nagging injuries that catch up to them in the postseason (Paul). But in general, guys who miss time in the regular season also miss time in the playoffs — there is a predictive nature to the larger regular season in that sense.

A more probabilistic-based approach to injuries is probably more fair. Maybe I should assume that a player misses some number of playoff games in a given season based on his injury history across adjacent seasons at that point in his career. Depending on style of play and the nature of the injury, that might mean retrospectively dinging someone for health, even though they were quite healthy until that point in their career. (eg Derrick Rose) I’m not sure how to implement this yet, but a probabilistic approach could remain objective while reducing the “can’t get injured in Cancun” problem. Perhaps this will be part of a CORP update in the future.

---

For now, it’s the same results-oriented approach to health. In the table below, you’ll find the latest valuations from Greatest Peaks along with:

• A base +/- estimate with offensive and defensive estimates. Think of this as performance on a slightly-below .500 team, which is the middle part of the bell curve. (Removing players from average teams makes them slightly below average.)

• A scalability curve (portability) for a player’s offense, based on the base estimate. There are five different shapes, with -2 curves representing a steeper decline in value, and +2 holding value more on better and better teams. (None of these 5 curves are as extreme as the Iverson/Draymond examples from above.)

• Health, based on their regular season and playoff games played (GP)

• CORP ranges

The ranges are the newest addition, and represent a low-end evaluation and a high end valuation. The high-end of the range is how I view a player’s offense and defense in the best reasonable light at the same time, so if ramping up offense hurts defense, that’s taken into account. All valuations are based off my CORP calculator, with minor mental curving if I don’t think the five generic scalability curves apply.

Finally, let’s take everything we’ve outlined and apply it to Jordan and LeBron. Excluding health, the CORP calculator is solving for a player’s title equity based on three key inputs:

• Base offensive value.

• The scalability curve associated with that base value.

• Base defensive value. (The calculator views defense as scaling equally well for all players, since defense generally scales well.)

LeBron’s 2009 has a base +/- valuation of +7.38 with a -2 scaling curve (steeper decline) on offense. Jordan’s 1990 is +7.13 with a neutral scaling curve (more gradual decline). In other words, I think LeBron is going to be slightly better helping an average team, but he doesn’t hold that value as well as Jordan does as his team quality improves. So at some point — maybe by the time we reach, say, a +3 team? — Jordan will start to have greater impact.

Remember, it’s impossible to describe generalized impact without a scalability curve. We could achieve a similar CORP value with a different combination of base values and scalability curves, which is why LeBron’s 2013 and Jordan’s 1991 have nearly identical CORP, despite having different base values.

Hey Proxy, falcolombardi -- Just to expand on this a bit, in his introduction to the Top 40 Careers he wrote back in 2016, he also talks about what skills are considered more or less portable, and how he gets from plus minus to championship odds: https://backpicks.com/2017/12/11/the-backpicks-goat-the-40-best-careers-in-nba-history/.

For scalability, the main idea is that off-ball skills usually tend to lose less value with better teammates than on-ball skills. He's said there's plus minus studies that support this, and while I've found some general studies on this, I haven't found a separate study for each skill independently.

What goes into the scalability rating? He said he does a large amount of manual film study / tracking to categorize how much of a player's offense comes from these skills, and uses those individual studies (plus more detailed play-by-play / tracking data for the recent players) to come up with the approximate scalability rating. The scalability calculator isn't public, but an example might be "okay this player passes X% of the time and has a passer rating of A, he tries iso scoring Y% of the time, he has a spacing rating of Z%, etc., so the most likely scalability rating is B." You can also see more of the detailed play-by-play studies in a variety of videos on YouTube (e.g. LeBron videos, Curry videos, the video on how teammates affect each other, etc.)... for these, he'll look at how various metrics like scoring and plus-minus change for a player when playing with different teammates who have different skill-profiles. It's still a bit nebulous, as we don't have the exact formula, but this at least tells you the type of inputs he uses to approximate scalability.

How do you get from plus minus estimates and scalability to championship odds? You can calculate the odds of winning a championship in an average year based on team SRS / team wins. And you can calculate approximately how much a player would change/improve an average teams' SRS/Wins based on their plus minus data. Of course, individual value changes based on fit (and fitting well on good teams would increase your championship odds more), so that's where scalability comes in. Put the two together, and you can get an estimate for championship odds.

And since the discussion is talking about how these CORP evaluations might change over time: the plus minus estimate might change as you get more data or look closer into the data. And the scalability score might change as he does more film analysis, and so gets a larger sample for estimating their skill tendencies (so e.g. if a player is more or less off-ball, more or less iso-dependent, etc., in a larger film review, that might change their scalability score).

[LA Bird]

Here are the (new) results for round 26

Winner: 20 Davis

There were 8 voters in this round: AEnigma, trelos6, Dutchball97, DraymondGold, trex_8063, OhayoKD, falcolombardi, capfan33

A total of 31 seasons received at least 1 vote: 03 McGrady, 05 Nash, 06 Nash, 07 Nash, 08 Paul, 09 Paul, 11 Howard, 13 Paul, 14 Paul, 15 Paul, 16 Paul, 18 Davis, 18 Harden, 19 Harden, 20 Davis, 20 Harden, 21 Embiid, 22 Embiid, 49 Mikan, 50 Mikan, 51 Mikan, 61 Baylor, 90 Barkley, 90 Ewing, 91 Barkley, 92 Malone, 93 Barkley, 93 Malone, 94 Ewing, 97 Malone, 98 Malone

Top 10 seasons: 20 Davis, 07 Nash, 08 Paul, 15 Paul, 50 Mikan, 06 Nash, 19 Harden, 51 Mikan, 05 Nash, 20 Harden

H2H record (1 season per player)
20 Davis: 0.630 (17-10)
07 Nash: 0.625 (15-9)
08 Paul: 0.458 (11-13)
50 Mikan: 0.400 (8-12)
19 Harden: 0.333 (7-14)

If voter turnout remains low in the next round, the voting window will be reduced to 48 hours to speed the project up.

[AEnigma]

Here are the results for round 26

Winner: 20 Davis

There were 7 voters in this round: AEnigma, trelos6, Dutchball97, trex_8063, OhayoKD, falcolombardi, capfan33

A total of 25 seasons received at least 1 vote: 03 McGrady, 05 Nash, 06 Nash, 07 Nash, 08 Paul, 09 Paul, 11 Howard, 13 Paul, 14 Paul, 15 Paul, 16 Paul, 18 Davis, 18 Harden, 19 Harden, 20 Davis, 20 Harden, 21 Embiid, 22 Embiid, 50 Mikan, 51 Mikan, 61 Baylor, 90 Barkley, 90 Ewing, 93 Barkley, 94 Ewing

Top 10 seasons: 20 Davis, 07 Nash, 08 Paul, 19 Harden, 15 Paul, 90 Ewing, 06 Nash, 20 Harden, 05 Nash, 13 Paul

H2H record (1 season per player)
20 Davis: 0.625 (15-9)
07 Nash: 0.619 (13-8)
08 Paul: 0.500 (11-11)
19 Harden: 0.381 (8-13)
90 Ewing: 0.364 (8-14)

If voter turnout remains low in the next round, the voting window will be reduced to 48 hours to speed the project up.

You missed DraymondGold’s vote (Mikan/Paul/Nash), which may affect Davis versus Nash.

[LA Bird]

Here are the results for round 26

Winner: 20 Davis

There were 7 voters in this round: AEnigma, trelos6, Dutchball97, trex_8063, OhayoKD, falcolombardi, capfan33

A total of 25 seasons received at least 1 vote: 03 McGrady, 05 Nash, 06 Nash, 07 Nash, 08 Paul, 09 Paul, 11 Howard, 13 Paul, 14 Paul, 15 Paul, 16 Paul, 18 Davis, 18 Harden, 19 Harden, 20 Davis, 20 Harden, 21 Embiid, 22 Embiid, 50 Mikan, 51 Mikan, 61 Baylor, 90 Barkley, 90 Ewing, 93 Barkley, 94 Ewing

Top 10 seasons: 20 Davis, 07 Nash, 08 Paul, 19 Harden, 15 Paul, 90 Ewing, 06 Nash, 20 Harden, 05 Nash, 13 Paul

H2H record (1 season per player)
20 Davis: 0.625 (15-9)
07 Nash: 0.619 (13-8)
08 Paul: 0.500 (11-11)
19 Harden: 0.381 (8-13)
90 Ewing: 0.364 (8-14)

If voter turnout remains low in the next round, the voting window will be reduced to 48 hours to speed the project up.

You missed DraymondGold’s vote (Mikan/Paul/Nash), which may affect Davis versus Nash.

Fixed. Thank you for pointing it out.