What's new
By:

Somewhat scientific predictions for Jazz players stats and total win number for upcoming season

Looking back over the past 6 drafts, 28 wins averages out to the 7-8th spot in the draft (before the lottery).
 
[size/HUGE] fixed [/size];637587 said:
That could get us Exum. I'd be happy.
Click to expand...

I'd be thrilled with the playoffs.
I'd be OK with a pick in the 11-14 range
I'd be happy with pick in the 6-10 range.
I'd be thrilled with a top 5 pick.
I'd be stoked with a top 3 pick.

I think we land in the 6-10 range this year, the core develops well, we land our 2014 picks and shore up the bench for next year. Then it is time to start some serious ballin.
 
Stoked said:
I'd be thrilled with the playoffs.
I'd be OK with a pick in the 11-14 range
I'd be happy with pick in the 6-10 range.
I'd be thrilled with a top 5 pick.
I'd be stoked with a top 3 pick.

I think we land in the 6-10 range this year, the core develops well, we land our 2014 picks and shore up the bench for next year. Then it is time to start some serious ballin.
Click to expand...

So basically if we don't get a top 3 pick you're changing your handle?
 
I think that number is too high. Simply because most teams coming to Utah in the first months on back to backs are playoff locks. There won't be many fatigue and altitude wins. It's based on win shares. "Support" players have usually (relative POV) better win share numbers than those getting double teamed and stuff. This year there's no Big Al and Paul on who the opponents focus. I think 28 is a stretch.
 
"Somewhat scientific."

Some college journalist took a shot in the dark at how many minutes each of our players will play next year, then compared them to p36 numbers from last season. :|
 
Xsy said:
"Somewhat scientific."

Some college journalist took a shot in the dark at how many minutes each of our players will play next year, then compared them to p36 numbers from last season. :|
Click to expand...

Do you have a "morewhat scientific" approach you would like to share with the class.
 
I agree with akmvp but I do think Kanter could be a wild card this year. I see him as the focus of the post offense over favors.

Nice to see win share 3's compiled like that.
 
Class starting now.....

heyhey said:
Do you have a "morewhat scientific" approach you would like to share with the class.
Click to expand...


Jared Bray simply summed all the winshares per 36 to come up with Jazz wins, which indicates that he does not understand the concept of winshares (he is implicitly giving equal weight to Hayward and Biendris).

Correctly applied:

For each player:
....Winshares per minute = (winshares per 36 min) divided by 36
....Winshares = (winshares per minute) times (minutes played per game)

For a team:
.....Team forecast wins = sum of all the winshares for each player on the team (from calculation above)

If you do this correctly (with his assumed playing time minutes) you wind up with the Jazz with 21 wins, not 28.

Class dismissed. : - )
 
silesian said:
Jared Bray simply summed all the winshares per 36 to come up with Jazz wins, which indicates that he does not understand the concept of winshares (he is implicitly giving equal weight to Hayward and Biendris).
Click to expand...

Bray made a mistake, but not the mistake you accuse him of (for example, Evans' projected WS/minute according to Basketball Reference is nearly 3 times Burks's WS/minute -- yet Bray shows them both at 2.2). He understood how to correctly factor in minutes played. He seems to have done everything correctly, except all the final calculations are low by 22% (all his numbers are 78% what they should be if he had calculated correctly -- I have no idea how he made this mistake). If he had calculated correctly, he would have had the Jazz at 36 wins.

You are correct basically in how you attempt to correct Bray, except you have misinterpreted Bray's numbers (as far as I can tell). The numbers he gives (for example, Alec Burks: 2.2) is not WS/36, but the total winshares the players will be responsible for at the end of the year. (In fact, both you and Bray misinterpret the Basketball Reference data. It's not WS/36, it's WS/48.)

Anyway, according to this "semi-scientific" approach (which is interesting in a certain way, but I think does have problems: for example the BR projections are too dependent on career average rather than career arc), the Jazz should win 36 games.

The bigger issue, if these predictions were to somehow come true is that our presumed bench has a higher WS/48 average than our presumed starters. In fact, our bench, if the basketball reference forecast is correct, will probably be one of the best in the league, as the average of its 7 players that go into the calculation is of almost exactly average NBA players (.096; average =.100). Our starters, on average, compute to a bit below average NBA players (.086).
 
idiot said:
Bray made a mistake, but not the mistake you accuse him of (for example, Evans' projected WS/minute according to Basketball Reference is nearly 3 times Burks's WS/minute -- yet Bray shows them both at 2.2). He understood how to correctly factor in minutes played. He seems to have done everything correctly, except all the final calculations are low by 22% (all his numbers are 78% what they should be if he had calculated correctly -- I have no idea how he made this mistake). If he had calculated correctly, he would have had the Jazz at 36 wins.

You are correct basically in how you attempt to correct Bray, except you have misinterpreted Bray's numbers (as far as I can tell). The numbers he gives (for example, Alec Burks: 2.2) is not WS/36, but the total winshares the players will be responsible for at the end of the year. (In fact, both you and Bray misinterpret the Basketball Reference data. It's not WS/36, it's WS/48.)

Anyway, according to this "semi-scientific" approach (which is interesting in a certain way, but I think does have problems: for example the BR projections are too dependent on career average rather than career arc), the Jazz should win 36 games.

The bigger issue, if these predictions were to somehow come true is that our presumed bench has a higher WS/48 average than our presumed starters. In fact, our bench, if the basketball reference forecast is correct, will probably be one of the best in the league, as the average of its 7 players that go into the calculation is of almost exactly average NBA players (.096; average =.100). Our starters, on average, compute to a bit below average NBA players (.086).
Click to expand...

Thanks for the feedback. Could you post the link where the 2013/14 projected WS/48 is located? I hunted and could not find it.
 
silesian said:
Thanks for the feedback. Could you post the link where the 2013/14 projected WS/48 is located? I hunted and could not find it.
Click to expand...

It took me a while to figure out -- perhaps the projections are a new feature of the Basketball Reference site. Just go into the pages of any of the players, and the very top statistical table is a 2013-14 projection. The WS/48 stat is the one on the far right of the table. For example, https://www.basketball-reference.com/players/h/haywago01.html, show's Hayward's projected WS/48 at .110.

I don't know how Bray estimated Burke and Clark, since they're rookies and don't have projections. But their projections worked out to be about .058 and .041 WS/48 respectively.
 
idiot said:
It took me a while to figure out -- perhaps the projections are a new feature of the Basketball Reference site. Just go into the pages of any of the players, and the very top statistical table is a 2013-14 projection. The WS/48 stat is the one on the far right of the table. For example, https://www.basketball-reference.com/players/h/haywago01.html, show's Hayward's projected WS/48 at .110.

I don't know how Bray estimated Burke and Clark, since they're rookies and don't have projections. But their projections worked out to be about .058 and .041 WS/48 respectively.
Click to expand...

I wanted to write "thanks, Idiot" but thought that might be misconstrued!
 
idiot said:
The bigger issue, if these predictions were to somehow come true is that our presumed bench has a higher WS/48 average than our presumed starters. In fact, our bench, if the basketball reference forecast is correct, will probably be one of the best in the league, as the average of its 7 players that go into the calculation is of almost exactly average NBA players (.096; average =.100). Our starters, on average, compute to a bit below average NBA players (.086).
Click to expand...
I've never put much stock in WS and this proves my skepticism.
 
You must log in or register to reply here.

Share this page
Back
Top