Predictions 2014-2015

[fpliii]

Speaking of b-r.com, for the last couple of years at least, they've had something like 'Projected Performance' (for the upcoming season) on player pages in the preseason. Seems one might have to sift thru every player page to gather these. Has anyone done that, who wants to say how it's done most practically, or whether it's worth much?

An 'aging' factor would be useful. But a 'bounceback' factor -- for younger players who've just had a down year -- would also be good. This Bucks team seems to have gathered a bunch of such players this season.

Are those b-r.com player projections still available somehow?

These:

http://www.basketball-reference.com/fri ... ctions.cgi

?

You can select any season, back to 80-81, from the dropdown menu at the top of the page.

[DSMok1]

Speaking of b-r.com, for the last couple of years at least, they've had something like 'Projected Performance' (for the upcoming season) on player pages in the preseason. Seems one might have to sift thru every player page to gather these. Has anyone done that, who wants to say how it's done most practically, or whether it's worth much?

An 'aging' factor would be useful. But a 'bounceback' factor -- for younger players who've just had a down year -- would also be good. This Bucks team seems to have gathered a bunch of such players this season.

Are those b-r.com player projections still available somehow?

They are here: http://www.basketball-reference.com/fri ... ctions.cgi

They are quite basic "Simple Projection System" projections. Dave Corby has been talking about revisiting the methodology, particularly for composite stats like BPM.

[J.E.]

Here's a chart showing the Bucks' cumulative point differential after each game

Jabari got injured in game #25

[Mike G]

That's very interesting. I see Larry Sanders has been out since G29, which is about when their season turned for the better.
Ilyasova has only played a handful of games since G25.
(I hadn't followed the Bucks, just now looking it up; obviously my comment that "they're 2 deep at every position" hasn't been especially true)

John Henson returned to action in G30, coinciding with the upswing.
One could look up other players' trends, and surely they show improvement since late Dec. Does the absence of certain players make others play better? Or is it just coincidence?

[J.E.]

Ah yes, I forgot about Sanders.

Here's everyone's +/- from season start until Game #25 (Dec 15th)

Code: Select all

╔═══════════════════════╦═══════════╗
║         Name          ║ PlusMinus ║
╠═══════════════════════╬═══════════╣
║ Kendall Marshall      ║ 9.9       ║
║ Jerryd Bayless        ║ 7.8       ║
║ Jared Dudley          ║ 6.5       ║
║ Khris Middleton       ║ 5.9       ║
║ Zaza Pachulia         ║ 2.9       ║
║ Ersan Ilyasova        ║ 0.4       ║
║ John Henson           ║ -0.9      ║
║ O.J. Mayo             ║ -3.2      ║
║ Giannis Antetokounmpo ║ -3.5      ║
║ Brandon Knight        ║ -7.1      ║
║ Jabari Parker         ║ -7.9      ║
║ Larry Sanders         ║ -8.5      ║
╚═══════════════════════╩═══════════╝

When the two worst players (in a +/- sense) on this list dropped out of rotation - in Parker's case unwillingly - the team's overall +/- started to make the rise that you see 2 posts above. Of course, it's not always that simple.

With raw +/-, one doesn't know for certain whether e.g. Sanders was dragging Parker's +/- down (or vice versa); though most metrics indicate that Parker was indeed a negative

[Mike G]

From Oct. 10 in this thread:

I like Milwaukee as a sleeper. Sanders and Ilyasova were IMO the best players on their playoff team 2 years ago and are still there even if in sub-optimal condition, I like the trio of Mayo, Middleton, Dudley shooting 3s at SG/SF, Giannis could have a good year, Jason Kidd's combination of spacing and aggressive defense seems perfect for making a team over perform. The other thing about Milwaukee is unless Parker sucks, they don't have any other scrub parts of their rotation, unlike other tanking teams. They have like 12-13 reasonable pros or better (Knight, Bayless, Marshall, Wolters, Mayo, Middleton, Dudley, Giannis, Ilyasova, Henson, Sanders, Pachulia + Parker)

Having 3-5 below-avg players on the floor at all times, they won't beat many good teams; but there are lots of weak teams in the East, so I guess they'll win 20-25. And that's figuring Jabari Parker goes 20 mpg at nearly-avg proficiency.

And now comes J.E.'s table showing all the early-season starters with terrible (raw) plus-minus. A couple of these starters go out, and the team turns it around.

The subs were better than most teams' subs. But this surely isn't typical behavior: You can't generally assume that by replacing your low-PM starters with better-PM subs, your team will be better.

The only consistent story line that I can see here is the head coach.
Last year, thru Dec. 31, the Nets were 10-21; went 12-4 thru Feb 6
This year's Bucks were 15-16 thru Dec. 27; 12-7 since.

ps: Dr Pos predicted 36 wins for Mil. That is 9.3 short of their current projection; the rest of us are off by 14 to 25 .

[Dr Positivity]

I was projecting Sanders and Ilyasova to be 2 of their 4 most valuable players (with Middleton and Dudley) so would have been farther off than that if I'd had a better finger on their pulse. If they had 2012-2013 Sanders/Ilyasova this would be over 50 W team?

[J.E.]

Some more info on Dudley's injury while with the Clippers http://espn.go.com/nba/story/_/id/12316 ... s-clippers
Apparently he played with a knee fracture (OUCH!) at Doc Rivers' request

Some of these coaches can still learn a thing or two from Gregg Popovich

[sndesai1]

the player tracking plus minus entry has taken a huge lead (might be insurmountable at this point)
vegas entry seems to be only getting worse in terms of ranking

[Mike G]

Holy crap. If this were Survivor, I'd be picking out my walking-the-plank attire.

Edit: Did anyone else use a single metric to make their predictions? With the objective of letting eWins do what it would do, I deliberately avoided any speculative adjustment based on team culture, player chemistry, coaching, etc. Obviously, these all have their place when you're trying to guess a team's prospects. They also take some time and effort.

[DSMok1]

Holy crap. If this were Survivor, I'd be picking out my walking-the-plank attire.

Edit: Did anyone else use a single metric to make their predictions? With the objective of letting eWins do what it would do, I deliberately avoided any speculative adjustment based on team culture, player chemistry, coaching, etc. Obviously, these all have their place when you're trying to guess a team's prospects. They also take some time and effort.

I think bbstats used purely RPM, and the leader of course used only PT-PM.

[AcrossTheCourt]

Holy crap. If this were Survivor, I'd be picking out my walking-the-plank attire.

Edit: Did anyone else use a single metric to make their predictions? With the objective of letting eWins do what it would do, I deliberately avoided any speculative adjustment based on team culture, player chemistry, coaching, etc. Obviously, these all have their place when you're trying to guess a team's prospects. They also take some time and effort.

I think bbstats used purely RPM, and the leader of course used only PT-PM.

From his site:
"This year I am using an exploratory player metric I created using publicly released SportVu data called Player Tracking Plus Minus (PT-PM) blended with RAPM and will follow up with another projection using ASPM Daniel Myers' boxscore based player metric."

What's interesting is how well he's doing and how his win predictions were so low: a bunch of teams in the west at 51 wins and only two in the mid-50's. What really helped is that his projections stayed away from the extremes.

[Crow]

The top 5 all have some measure of RAPM in them.


Schoene stopped competing and BR simple projections never did. Wonder how BPM would look. Better than ASPM I assume. Has Arturo commented on the performance of his WP projections? Haven't looked in that direction recently.

[bbstats]

I used RPM projections yep. FiveThirtyEight used RPM as well.

Only non-one-number-adjustment I used was a lazy one for Usage%.

[bbstats]

Holy crap. If this were Survivor, I'd be picking out my walking-the-plank attire.
What's interesting is how well he's doing and how his win predictions were so low: a bunch of teams in the west at 51 wins and only two in the mid-50's. What really helped is that his projections stayed away from the extremes.

I wonder if injuries are partially to credit here. KD/LeBron/Kawhi being out regressed the top teams a lot.