Author Topic: Celtics predicted to have fewer wins?!?!  (Read 3298 times)

0 Members and 1 Guest are viewing this topic.

Celtics predicted to have fewer wins?!?!
« on: October 20, 2016, 04:16:30 PM »

Offline The One

  • Bailey Howell
  • **
  • Posts: 2017
  • Tommy Points: 203
http://projects.fivethirtyeight.com/2017-nba-predictions/?addata=espn:frontpage

FiveThirtyEight.com's CARMELO is projecting the Celtics to have 47 wins this year...one less than last year!?!?

That doesn't make sense to me...added Horford and Brown.  Smart, Rozier, and Bradley look like they have improved.

Those factors to me show that the Celtics will win 53 (my official prediction.

Do you guys see the team having fewer wins??

Can someone please explain?

(I do like seeing the Nets predicted to have the worst record in the NBA... ;D ;D ;D)

Re: Celtics predicted to have fewer wins?!?!
« Reply #1 on: October 20, 2016, 04:32:26 PM »

Offline PhoSita

  • NCE
  • Robert Parish
  • *********************
  • Posts: 21835
  • Tommy Points: 2182
I think it's important to remember that the 538 projection is the average of simulated outcomes. 

Perhaps the 538 model looks at the Celts 48 wins last year and thinks that, all things considered, that was closer to a best case outcome for that team.  Perhaps it also thinks that the loss of Turner and Sullinger is significant enough to substantially eat into any potential improvement attributable to the addition of Horford, Brown, and Green.

The range of significantly likely outcomes for the Celts in the eyes of their model could be, say, 44 to 55 wins, with the median skewing toward the lower end of that range.

Personally I think that 48 wins was a very good outcome for last year's team.  At the same time, I think they're now head and shoulders above all the other Eastern Conference teams outside of Toronto and Cleveland, so winning less than 50 games would be an underachievement as far as I'm concerned.
You値l have to excuse my lengthiness葉he reason I dread writing letters is because I am so apt to get to slinging wisdom & forget to let up. Thus much precious time is lost.
- Mark Twain

Re: Celtics predicted to have fewer wins?!?!
« Reply #2 on: October 20, 2016, 04:38:03 PM »

Offline PhoSita

  • NCE
  • Robert Parish
  • *********************
  • Posts: 21835
  • Tommy Points: 2182
The 538 predictions are pretty wacky this year though.

They have the Thunder winning more games than the Clippers and Raptors, and the Bulls winning 45 with an 80% chance at the playoffs.
You値l have to excuse my lengthiness葉he reason I dread writing letters is because I am so apt to get to slinging wisdom & forget to let up. Thus much precious time is lost.
- Mark Twain

Re: Celtics predicted to have fewer wins?!?!
« Reply #3 on: October 20, 2016, 04:43:40 PM »

Online Who

  • James Naismith
  • *********************************
  • Posts: 47135
  • Tommy Points: 2401
I would guess that (1) they expect Horford to slow down / decline some (2) Jaylen Brown rates quite poorly given his poor scoring inefficiency in college and is expected to hurt rather than help the team as a rookie against the loss of the productive Evan Turner from last year's team. 

Re: Celtics predicted to have fewer wins?!?!
« Reply #4 on: October 20, 2016, 04:45:04 PM »

Offline fairweatherfan

  • Johnny Most
  • ********************
  • Posts: 20738
  • Tommy Points: 2365
  • Be the posts you wish to see in the world.
I'm too lazy/busy to look but does anyone have last year's predictions?

Re: Celtics predicted to have fewer wins?!?!
« Reply #5 on: October 20, 2016, 04:47:15 PM »

Offline kozlodoev

  • NCE
  • Kevin Garnett
  • *****************
  • Posts: 17914
  • Tommy Points: 1294
I think it's important to remember that the 538 projection is the average of simulated outcomes. 

Perhaps the 538 model looks at the Celts 48 wins last year and thinks that, all things considered, that was closer to a best case outcome for that team.  Perhaps it also thinks that the loss of Turner and Sullinger is significant enough to substantially eat into any potential improvement attributable to the addition of Horford, Brown, and Green.

The range of significantly likely outcomes for the Celts in the eyes of their model could be, say, 44 to 55 wins, with the median skewing toward the lower end of that range.

Personally I think that 48 wins was a very good outcome for last year's team.  At the same time, I think they're now head and shoulders above all the other Eastern Conference teams outside of Toronto and Cleveland, so winning less than 50 games would be an underachievement as far as I'm concerned.
Coincidentally, these are the only teams projected to win more than the Celtics. No-one else is expected to win more than 45.
"I don't know half of you half as well as I should like; and I like less than half of you half as well as you deserve."

Re: Celtics predicted to have fewer wins?!?!
« Reply #6 on: October 20, 2016, 04:50:18 PM »

Offline saltlover

  • Frank Ramsey
  • ************
  • Posts: 12490
  • Tommy Points: 2619
I'm not surprised.  If you looked at the CARMELO projections earlier this summer, most of the Celtics players were projected to have a decrease in WAR this year, as last year most of them posted huge spikes in that stat.  The spikes were due in part to an increase in playing time for many players, be it due to role (Crowder) or better health (Bradley).  The projection system assumes that many of these players will play fewer minutes, and accordingly lowers the WAR.  I would assume the win total is largely based on just summing WAR's together, but even if more robust, CARMELO forecasted declines in part for this reason.

Re: Celtics predicted to have fewer wins?!?!
« Reply #7 on: October 20, 2016, 04:57:51 PM »

Offline alldaboston

  • Antoine Walker
  • ****
  • Posts: 4170
  • Tommy Points: 324
It's an issue with 538 and ELO/CAREMELO

I quote, from a Reddit user:

Quote
And this year they took RPM out of the modeling and are exclusively using ELO and BPM. I strongly disagree with that decision and it looks like the model spit out some weird stuff this year. Even the guy that created BPM defers to other projections with APMs blended in. They made CARMELO worse for better player comparisons, but I think it had a brutal effect on their overall metric.

https://www.reddit.com/r/nba/comments/58i38e/fivethirtyeight_538s_20162017_nba_season/d90s79j
I could very well see the Hawks... starting Taurean Prince at the 3, who is already better than Crowder, imo.

you vs. the guy she tells you not to worry about

Re: Celtics predicted to have fewer wins?!?!
« Reply #8 on: October 20, 2016, 05:12:04 PM »

Offline Chris22

  • Rajon Rondo
  • *****
  • Posts: 5081
  • Tommy Points: 460
It all depends on injuries, and we already lost Olynyk and Smart.

Re: Celtics predicted to have fewer wins?!?!
« Reply #9 on: October 20, 2016, 05:58:27 PM »

Offline The Fawb

  • Oshae Brissett
  • Posts: 50
  • Tommy Points: 3
I think ELO, BPM, APM and carmelo are all terrible. Not one can stand up to regression analysis  and all have their stat value constants picked with no scientific or statistical means, they actually just pick whatever number they feel is good. That's why Wins Produced is so much better. It is derived from a regression analysis for each box score stat, based of its value in wins. Much more accurate. It is a descriptive stat, not predictive,  but can be used fairly well in that way. Www.boxscoregeeks.com

Re: Celtics predicted to have fewer wins?!?!
« Reply #10 on: October 20, 2016, 06:35:21 PM »

Offline PhoSita

  • NCE
  • Robert Parish
  • *********************
  • Posts: 21835
  • Tommy Points: 2182
I think it's important to remember that the 538 projection is the average of simulated outcomes. 

Perhaps the 538 model looks at the Celts 48 wins last year and thinks that, all things considered, that was closer to a best case outcome for that team.  Perhaps it also thinks that the loss of Turner and Sullinger is significant enough to substantially eat into any potential improvement attributable to the addition of Horford, Brown, and Green.

The range of significantly likely outcomes for the Celts in the eyes of their model could be, say, 44 to 55 wins, with the median skewing toward the lower end of that range.

Personally I think that 48 wins was a very good outcome for last year's team.  At the same time, I think they're now head and shoulders above all the other Eastern Conference teams outside of Toronto and Cleveland, so winning less than 50 games would be an underachievement as far as I'm concerned.
Coincidentally, these are the only teams projected to win more than the Celtics. No-one else is expected to win more than 45.


That's a good point.

Win totals overall in a model like this are somewhat deflated because it pins every team at their "average" outcome.

Yet we know that each year there are a number of teams that suffer injuries, deal with chemistry issues and off-court distractions, and end up out of contention by January, leading to a steep downward trajectory the rest of the way.

There are also teams each year that exceed expectations, go on extended hot streaks, and so on.



Over the last 6-7 years, the top 2 seeds in the East have won:

56 & 48
53 & 50
54 & 48
54 & 49
57 & 52 (adj for lockout)
58 & 56
59 & 53


Basically, if you're a top 3 seed, chances are you're going to win 50-55 games.
You値l have to excuse my lengthiness葉he reason I dread writing letters is because I am so apt to get to slinging wisdom & forget to let up. Thus much precious time is lost.
- Mark Twain

Re: Celtics predicted to have fewer wins?!?!
« Reply #11 on: October 20, 2016, 06:36:47 PM »

Offline PhoSita

  • NCE
  • Robert Parish
  • *********************
  • Posts: 21835
  • Tommy Points: 2182
I'm too lazy/busy to look but does anyone have last year's predictions?

I think it was pretty bullish on the Celts last year.  Had them for 47 or 48, which of course turned out to be accurate.
You値l have to excuse my lengthiness葉he reason I dread writing letters is because I am so apt to get to slinging wisdom & forget to let up. Thus much precious time is lost.
- Mark Twain

Re: Celtics predicted to have fewer wins?!?!
« Reply #12 on: October 20, 2016, 07:39:32 PM »

Offline The One

  • Bailey Howell
  • **
  • Posts: 2017
  • Tommy Points: 203
I think ELO, BPM, APM and carmelo are all terrible. Not one can stand up to regression analysis  and all have their stat value constants picked with no scientific or statistical means, they actually just pick whatever number they feel is good. That's why Wins Produced is so much better. It is derived from a regression analysis for each box score stat, based of its value in wins. Much more accurate. It is a descriptive stat, not predictive,  but can be used fairly well in that way. Www.boxscoregeeks.com

So using the method you describe...what would the prediction be for this year's Celtics?

Re: Celtics predicted to have fewer wins?!?!
« Reply #13 on: October 20, 2016, 08:46:17 PM »

Offline GreenWarrior

  • Ray Allen
  • ***
  • Posts: 3275
  • Tommy Points: 228
I think Jackie Macmullan and me are the only ones that think the loss of turner is huge. that's not to say I don't think we could or should be better because we did add another all-star.

but I absolutely wouldn't be surprised if we were at or around the same amount of wins as last yr. either.

I think the "emergence" of rozier is way over-blown(not surprising on this board). he will have growing pains and I think he will struggle with the one thing turner excelled at and that's if his shot aint falling impact the game in other ways. that and pretty much the overall chemistry on the bench will boil down to weather rozier had a good shooting night imo. because he's not the creator that turner was either.

Re: Celtics predicted to have fewer wins?!?!
« Reply #14 on: October 20, 2016, 08:52:27 PM »

Offline Moranis

  • James Naismith
  • *********************************
  • Posts: 33461
  • Tommy Points: 1533
I've said it before on here and will say it again, while Horford is a significantly better player than Sullinger, Horford's strengths were already strengths of the team and his weaknesses are still weaknesses on the team, and the one area that Sullinger was better than Horford was rebounding.  Boston projects to be a pretty poor rebounding team this year.  Boston still has no shot blocker.  Still has no #1 wing scoring option and only Thomas as a real #1 scoring option and he is so small he can be neutralized quite easily in the playoffs.  Boston still has poor and inconsistent 3 point shooting.  The loss of Turner also can't be understated as only Turner and Thomas last year were any good at shot creation (and that is not a strength of Horford either).  Now maybe Rozier can continue his good play from the preseason or maybe Smart makes this miraculous leap offensively, but I don't think any of those are very likely. 

You also can't discount just how many games Boston had against terrible teams.  Philadelphia, Brooklyn, and New York were the three worst teams in the Eastern Conference last year, all of them were in the Atlantic.  And while Philly and Brooklyn will still be poor teams, I would expect them to be better than the 31 combined wins they had.  New York should be a much better team this year. 
2023 Historical Draft - Brooklyn Nets - 9th pick

Bigs - Pau, Amar'e, Issel, McGinnis, Roundfield
Wings - Dantley, Bowen, J. Jackson
Guards - Cheeks, Petrovic, Buse, Rip