Overall Tracking of Player performances

[Yit]

Okay, we've cross the 25 games mark, so a quick review of what I have been tracking for our players. It is what I call a Game Contribution Point (GCP) score. This in short is a computation of basic box score stats into a single value. Essentially, it takes a few things into consideration (e.g. time normalization, shooting normalization etc.).

Overall GCP for our players +/- Standard Deviation (n=number of valid games scored)
Lopez 36 +/- 15 (23)
Lin 28 +/- 8 (7)
Booker 26 +/- 13 (24)
Kilpatrick 26 +/- 15 (25)
Bogdanovic 23 +/- 13 (25)
Harris 15 +/- 8 (25)
RHJ 15 +/- 11 (22)
Hamilton 13 +/- 11 (21)
Scola 13 +/- 12 (21)
Whitehead 11 +/- 6 (21)
Bennett 11 +/- 7 (18)
Levert 9 +/- 5 (5)
Foye 8 +/- 6 (17)
Ferrell 8 +/- 8 (10)
McCullough 7 +/- 11 (9)
Dinwiddie 4 +/- 9 (4)

In general, we expect a normal NBA player to play around 10 - 19 range, a good NBA player would be playing above an average GCP of 20. The drop from our best 5 players to the next tier is quite drastic, ideally we should try to grow some players towards the 17-19 range. Rookies in development would be averaging below 10, Whitehead is starting to break into the normal range which is a clear sign that his development is on track. The stats for Dinwiddie isn't quite useful at the moment as the sample size is too small, we will see how it goes after a few more games.
So what does it all say about our team, essentially the only star player (playing above 30 GCP) is Lopez (nothing new). Lin isn't quite there yet, so hopefully when he has recovered fully and perform better he should be knocking into the 30s. Overall, the Nets is a normal team - we can't expect too much out of this year's squad, but it would be great to see the team growing in strength and ability. Will see how we fare by the half-way mark (around 41 games mark) but it is nice to see the team continuing to play hard. Go Nets!

[leeramundo]

Can you tell us or message me what the actual calculation is?

[Yit]

Can you tell us or message me what the actual calculation is?

Okay for those interested,

Basically, we take the following into consideration:
There are 3 components to the GCP = Score Index + Bonus - Penalties + (+/- factor impact)

Score Index is computed by Points Scored + Shooting Efficiency Modifications + Time moderation modifications
Shooting Efficiency Modifications = (PlayerShooting%-TeamShooting%)*PS (note this can be either positive or negative - main purpose is to mitigate effect of volume shooting)
Points Scored + SEM = BaseScore
Time moderation is achieved by = ((BaseScore/PlayingTime)-(TeamScore/TotalTeamPlayingTime))*BaseScore)

Final Score Index calculated would be BS + Time moderation.

Bonuses are straight forward addition of Total Rebounds + Assists + Steals + Blocks

Penalties are TO and Blocked Shots

+/- Factor (which can be positive or negative) = +/- score/5 (assuming any +/- contribution is by 5 players playing together as such a player is only getting 20% of the deal done - note this can either be positive or negative).

Adding them all up and rounding to full number.

Hope I don't confuse anybody (I am still considering whether need to tweak certain weightage but so far I think it holds out reasonably well).

[MrDollarBills]

Okay, we've cross the 25 games mark, so a quick review of what I have been tracking for our players. It is what I call a Game Contribution Point (GCP) score. This in short is a computation of basic box score stats into a single value. Essentially, it takes a few things into consideration (e.g. time normalization, shooting normalization etc.).

Overall GCP for our players +/- Standard Deviation (n=number of valid games scored)
Lopez 36 +/- 15 (23)
Lin 28 +/- 8 (7)
Booker 26 +/- 13 (24)
Kilpatrick 26 +/- 15 (25)
Bogdanovic 23 +/- 13 (25)
Harris 15 +/- 8 (25)
RHJ 15 +/- 11 (22)
Hamilton 13 +/- 11 (21)
Scola 13 +/- 12 (21)
Whitehead 11 +/- 6 (21)
Bennett 11 +/- 7 (18)
Levert 9 +/- 5 (5)
Foye 8 +/- 6 (17)
Ferrell 8 +/- 8 (10)
McCullough 7 +/- 11 (9)
Dinwiddie 4 +/- 9 (4)

In general, we expect a normal NBA player to play around 10 - 19 range, a good NBA player would be playing above an average GCP of 20. The drop from our best 5 players to the next tier is quite drastic, ideally we should try to grow some players towards the 17-19 range. Rookies in development would be averaging below 10, Whitehead is starting to break into the normal range which is a clear sign that his development is on track. The stats for Dinwiddie isn't quite useful at the moment as the sample size is too small, we will see how it goes after a few more games.
So what does it all say about our team, essentially the only star player (playing above 30 GCP) is Lopez (nothing new). Lin isn't quite there yet, so hopefully when he has recovered fully and perform better he should be knocking into the 30s. Overall, the Nets is a normal team - we can't expect too much out of this year's squad, but it would be great to see the team growing in strength and ability. Will see how we fare by the half-way mark (around 41 games mark) but it is nice to see the team continuing to play hard. Go Nets!

Actually just judging from the ratings this matches the eyeball test. Can you list a ratings chart like they do for PER? For exampl:

All-time great season 43+

Hands-down MVP 33-42

Strong MVP candidate 29.9-32.9

Long-shot MVP candidate 25-27.5

Definite All-Star 22.5-25

Borderline All-Star 20-22.5

Second offensive option 18-20

Third offensive option 16.5-18

Slightly above-average player 15-16.5

Rotation player 13-15

Non-rotation player 11-13

Fringe roster player 9-11

Player who won't stick in the league 0-8.7

[Yit]

Actually just judging from the ratings this matches the eyeball test. Can you list a ratings chart like they do for PER?

Not quite there yet, will continue to track first - but currently roughly breaking the players into 4 broad categories

Development players - GCP average below 10 (seriously if the player isn't a rookie and consistently only score this GCP they should really re-think their careers in NBA).
Normal players (or role players) - GCP average 10 to 19
Good players (starters) - GCP average 20 - 29
Star players - GCP average 30 and above.

[MrDollarBills]

thanks

[Yit]

Okay folks we have hit the ASW.

Time for a quick review - had been busy so am sort of relegated to analyzing during my weekends.

But using the Game Contribution Points (GCP) that I had been computing, I noticed something rather interesting.

So far the Nets have been averaging a Best-5-Player-Average GCP of around 28 points

However, our opponents on average have been doing B5PAGCP of about 36 points against us.

I have observed also that a differential of at least 3 or more points makes a sure win against opponents. What this means is that Nets will have to average around 34 points to have some kind of chance at beating opponents - i.e. to have a chance of having a 0.5 record. Even if we assume that we get lucky and Lin comes back - at most he might improve our average to say around 30 points, we are probably still 2 good players short of hitting the average mark.

General observations (still watching out) - it seems to me that Contenders (team likely to fight for the championship) should be averaging a B5PAGCP of above 40 points, and the teams that are likely to make playoffs will be playing at B5PAGCP of around 36 points and above.

In short, assuming playing at a B5PAGCP of 36 - gives you a chance of 0.5,
playing at 34 will give you a chance of 0.33
playing at 32 will give you a chance of 0.25
playing at 30 will give you a chance of 0.2
playing at 28 will give you a chance of 0.17 (which is where we are at the moment Nets is at 9/55 or around 0.16)

Okay - my predictions, is if Lin can come back we probably only hit the next tier (or 0.2) or giving us 16 wins....

[Yit]

Okay we have hit the 80 games mark, will be analysing our players' GCP over the season and proposed who we want to consider keeping or moving for the next season. Haven't got the energy to analyse the GCP of opposing teams (much - at most only 3 to 4 games i.e. when they play us - if I have more data might even be able to propose whom we should target or trade for).

But here's a general breakdown so far:

Overall GCP for our players +/- Standard Deviation (n=number of valid games scored)
Lopez 36 +/- 15 (74)
Lin 25 +/- 11 (35)
Booker 23 +/- 11 (71)
Kilpat 21 +/- 14 (70)
RHJ 18 +/- 11 (76)
CLV 16 +/- 11 (55)
Dinwiddie 15 +/- 11 (57)
Goodwin 15 +/- 13 (10)
Hamilton 14 +/- 10 (61)
Acy 14 +/- 11 (22)
Whitehead 13 +/- 8 (71)
Harris 13 +/- 9 (52)
KJM 11 +/- 10 (18)
Foye 9 +/- 8 (67)
Nicholson 7 +/- 8 (8)

If we assume that we replaced the bottom 5 - so that we can upgrade the team overall, at risk would be Nicholson, Foye, KJM, Harris and Whitehead (but Whitehead may still be retained as his is a rookie contract).

But seriously though, if we want to make the playoffs, we would need to add 1 or 2 more pieces at the top 5 - Players who can give us GCPs in the range of 36 and above.

Currently our Best 5 Players average (GCP) from 80 games is hovering around 28 - which is just enough to bring us to around 20.5 wins. In reality our players have already over-performed (if we compute our players average best 5 from table above, it is around 25).

[Yit]

Another use of GCP is to compute the economic value of our players i.e. are we getting value for our money, dividing their salaries by GCP

GCP +/- SD (n) Salary S/GCP
Lopez 36 +/- 15 (74) $21,165,675 $587,935 per GCP
Lin 25 +/- 11 (35) $11,483,254 $459,330 per GCP
Booker 23 +/- 11 (71) $9,250,000 $402,174 per GCP
Kilpat 21 +/- 14 (70) $980,431 $46,687 per GCP
RHJ 18 +/- 11 (76) $1,395,600 $77,533 per GCP
CLV 16 +/- 11 (55) $1,562,280 $97,643 per GCP
Dinwiddie 15 +/- 11 (57) $726,672 $48,445 per GCP
Goodwin 15 +/- 13 (10) $57,672 $3,845 per GCP
Hamilton 14 +/- 10 (61) $3,000,000 $214,286 per GCP
Acy 14 +/- 11 (22) $1,790,902 $127,926 per GCP
Whitehead 13 +/- 8 (71) $1,074,145 $86,627 per GCP
Harris 13 +/- 9 (52) $980,431 $75,418 per GCP
KJM 11 +/- 10 (18) $3,333,333 $303,030 per GCP
Foye 9 +/- 8 (67) $2,500,000 $277,778 per GCP
Nicholson 7 +/- 8 (8) $6,088,993 $869,856 per GCP

From a VFM perspective, Nicholson does not look like worth retaining, Goodwin looks like a steal and would be paid more if we offer a contract. Overall, our front office seems to maintain a good rate, although Kilpatrick is probably due a raise.

[TheNetsFan]

It would help if you could cherry-pick a variety of well known players and various positions & provide their scores as well. It seems to fairly accurately assess the players' relative impact within the team, but if 10-19 is normal, and that range includes guys like Goodwin, Dinwiddie, Whitehead, etc., what does a legit starter like Teague or a star like Paul get?

[wonchobody]

This is great stuff. Will there be a time where you will compute the ratings for upcoming free agents? This way we can see where ratings and potential contract values can be compared?

[Yit]

It would help if you could cherry-pick a variety of well known players and various positions & provide their scores as well. It seems to fairly accurately assess the players' relative impact within the team, but if 10-19 is normal, and that range includes guys like Goodwin, Dinwiddie, Whitehead, etc., what does a legit starter like Teague or a star like Paul get?

I don't have high number of games just the ones they played against us but for your interest

Jeff Teague GCP 10 + 29 + 58 + 52 or an average GCP of 37 over 4 games

Chris Paul GCP 59 + 58 or an average of 59 over 2 games against us.

Star level players in general would average 30 and above e.g. Teague
All-star level players in general would be averaging 40 and above in terms of GCP e.g. Paul

[Yit]

This is great stuff. Will there be a time where you will compute the ratings for upcoming free agents? This way we can see where ratings and potential contract values can be compared?

I can't promise anything - I have my day job to take care of. Ideally I may need to take a look at their latest 8 to 10 game stats to get a good gauge. Do you have any specifc player you are interested in - maybe I can try to get some computations squeezed in for some of them.

[2k15]

It would help if you could cherry-pick a variety of well known players and various positions & provide their scores as well. It seems to fairly accurately assess the players' relative impact within the team, but if 10-19 is normal, and that range includes guys like Goodwin, Dinwiddie, Whitehead, etc., what does a legit starter like Teague or a star like Paul get?

I don't have high number of games just the ones they played against us but for your interest

Jeff Teague GCP 10 + 29 + 58 + 52 or an average GCP of 37 over 4 games

Chris Paul GCP 59 + 58 or an average of 59 over 2 games against us.

Star level players in general would average 30 and above e.g. Teague
All-star level players in general would be averaging 40 and above in terms of GCP e.g. Paul

This doesn't provide a good gauge in my opinion. You'd need to normalize their score over their games against other opponents for the right context. Because they score 37 and 59 against the worst team in the NBA doesn't mean that's a useful score to measure them.

[wonchobody]

This is great stuff. Will there be a time where you will compute the ratings for upcoming free agents? This way we can see where ratings and potential contract values can be compared?

I can't promise anything - I have my day job to take care of. Ideally I may need to take a look at their latest 8 to 10 game stats to get a good gauge. Do you have any specifc player you are interested in - maybe I can try to get some computations squeezed in for some of them.

I see. So you have to go through each box score normalized for time? I can see how that would be time consuming. Would be cool to see how our Free Agent targets compare to the players they are replacing to get a rough gauge on how the team improves with a potential signing/addition.

So if possible, seeing players we may target like Otto Porter, KCP, Reddick, George Hill and potential value signings like Jamychal Green, Patty Mills, or someone of their ilk.

[Yit]

It would help if you could cherry-pick a variety of well known players and various positions & provide their scores as well. It seems to fairly accurately assess the players' relative impact within the team, but if 10-19 is normal, and that range includes guys like Goodwin, Dinwiddie, Whitehead, etc., what does a legit starter like Teague or a star like Paul get?

I don't have high number of games just the ones they played against us but for your interest

Jeff Teague GCP 10 + 29 + 58 + 52 or an average GCP of 37 over 4 games

Chris Paul GCP 59 + 58 or an average of 59 over 2 games against us.

Star level players in general would average 30 and above e.g. Teague
All-star level players in general would be averaging 40 and above in terms of GCP e.g. Paul

This doesn't provide a good gauge in my opinion. You'd need to normalize their score over their games against other opponents for the right context. Because they score 37 and 59 against the worst team in the NBA doesn't mean that's a useful score to measure them.

Ofcourse a 2-game sample won't be a good gauge - but just for interest sake - I have computed Chris Paul's last 10 games (most recent) for LAC stretching from 20 Mar to 8 Apr 2017.
His GCP was 37-55-30-18-38-54-56-58-46-45 - which averages out to be around 44 +/- 13.
He is still All-Star level as far as I am concern.

[Yit]

This is great stuff. Will there be a time where you will compute the ratings for upcoming free agents? This way we can see where ratings and potential contract values can be compared?

I can't promise anything - I have my day job to take care of. Ideally I may need to take a look at their latest 8 to 10 game stats to get a good gauge. Do you have any specifc player you are interested in - maybe I can try to get some computations squeezed in for some of them.

I see. So you have to go through each box score normalized for time? I can see how that would be time consuming. Would be cool to see how our Free Agent targets compare to the players they are replacing to get a rough gauge on how the team improves with a potential signing/addition.

So if possible, seeing players we may target like Otto Porter, KCP, Reddick, George Hill and potential value signings like Jamychal Green, Patty Mills, or someone of their ilk.

It may take me a while but for what it is worth here's my computation for George Hill (most recent 10 games) from the period of 11 Mar to 10 Apr 2017 (he missed games from 29 Mar to 8 Apr) - his 10 game GCP was 9-37-30-11-36-26-24-1-26-54. This gives him an average of 25 +/- 16 (not that much different from Jeremy Lin's GCP though)

[Yit]

This is great stuff. Will there be a time where you will compute the ratings for upcoming free agents? This way we can see where ratings and potential contract values can be compared?

I can't promise anything - I have my day job to take care of. Ideally I may need to take a look at their latest 8 to 10 game stats to get a good gauge. Do you have any specifc player you are interested in - maybe I can try to get some computations squeezed in for some of them.

I see. So you have to go through each box score normalized for time? I can see how that would be time consuming. Would be cool to see how our Free Agent targets compare to the players they are replacing to get a rough gauge on how the team improves with a potential signing/addition.

So if possible, seeing players we may target like Otto Porter, KCP, Reddick, George Hill and potential value signings like Jamychal Green, Patty Mills, or someone of their ilk.

Probably tackle one player at a time, here are the latest GCP for Otto Porter (recent 10 games average):

Games from 22 Mar to 12 Apr = 16-11-23-33-7-11-24-16-17-20, or average GCP of 18 +/- 7