What's new
By:
Filters

Study done on whether preseason success is a predictor of regular season success

S

svenguru

Active Member
https://www.82games.com/preseason.htm

Summed up this study basically finds that if a team sucked the prior year yet has a good preseason record that there is a high likelihood they will experience a big improvement in wins during the regular season. Teams that had an average of 25 wins the prior year and then had a winning record the following preaseason saw on average a gain of 19 victories during the regular season!
 
I have no idea whether or not this pre-season will translate into regular season success but I will say this pre-season reminds me A LOT of the 2003 pre-season after Stockton/Malone left (IIRC Jazz only lost one game that pre-season).

I think it all revolves around the fact that QS has made it readily known from the start that there will be minutes available to people who come to play - much like Sloan did when The Jazz had AK47, Harpring and about 15 other guys lucky to be in the league.
 
candrew said:
I have no idea whether or not this pre-season will translate into regular season success but I will say this pre-season reminds me A LOT of the 2003 pre-season after Stockton/Malone left (IIRC Jazz only lost one game that pre-season).

I think it all revolves around the fact that QS has made it readily known from the start that there will be minutes available to people who come to play - much like Sloan did when The Jazz had AK47, Harpring and about 15 other guys lucky to be in the league.
Click to expand...
Good observation. Guys are fighting for minutes and contracts. And Quin has let it be known it's not just about scoring points. I think the Jazz overachieve.
 
svenguru said:
https://www.82games.com/preseason.htm

Summed up this study basically finds that if a team sucked the prior year yet has a good preseason record that there is a high likelihood they will experience a big improvement in wins during the regular season. Teams that had an average of 25 wins the prior year and then had a winning record the following preaseason saw on average a gain of 19 victories during the regular season!
Click to expand...

Interesting find, even if outdated.

To me the single most important stat from the article was this: if a team is coming off of a 25 win (or worse) season, then the preseason record vs. regular season record has a correlation coefficient of 0.69. That's fairly significant.
 
How much is that correlation due to influx of a high, ready to play, draft pick?

Jazz don't have that.

What's the correlation to coaching changes?
 
Not sure if other teams play a more balanced preseason, but Utah plays 6 of their 8 against only 3 teams. Maybe the preseason is more indicative of the goals of each team.

As for the "high, ready-to-play pick," wasn't there a thread showing that the vast majority of rookies, even high picks, don't impact wins and losses as much as we think they might? I know colton sees a significant correlation, but I'm less convinced. I think it more useful to look at each team individually. For example, based on roster change, I see less chance of Philly improving than I do Milwaukee. And I know I'm a homer, but I like the Jazz' roster a whole lot better than Boston's, even if/when Rondo returns. Of course, there's the whole argument about WC vs. EC. So maybe strength of conference, since the schedule is unbalanced, also needs to be plugged into the equation.
 
Last edited:
All the caveats you all have noted apply to this data.

I'm also going to burst the bubble a bit:

As noted this is an old study with a small sample size. I just did the same study based on the most recent data (2009-10 season until now). Interestingly the two studies have the same number of teams with winning records in the preseason after losing more than 30 games (8 teams).

For this more recent data, the average winning percentage the following year of a team that had won less than 30 or more is virtually the same no matter what the following preseason winning percentage is. For all teams that had won less than 30 games the preceding year, the average number of wins in 30.89. For those that had an over .500 winning percentage in the preseason, the average number of wins is 31.775. So the difference is less than one game. Preseason winning percentages, at least recently, aren't really telling us anything.

In fact, the correlation (for those teams that had won less than 30 games) between the previous year's regular season record and the next year's percentage (0.46) is much stronger since 2009 than the correlation between the preseason and the regular season (0.10). (It's possible the lockout year with only two preseason games might be messing with that last correlation, however.)

So for every Charlotte and Phoenix from last year (previously bad teams that did well in the next preseason and regular season), there's a New Orleans and Sacramento (previously bad teams whose good following preseason didn't turn into a good regular season).
 
4 of the top 6 preseason teams (Cleveland, Brooklyn, Utah, Golden State) have new coaches
 
idiot said:
I'm also going to burst the bubble a bit:
Click to expand...

PS to my last post:

In examining more data (from the 2001-2 regular season until now), it's clear that that 82games result was the product of a either a small sample size or a real relationship that used to exist but no longer does. The results certainly don't apply recently.

In fact, if you break the results up from 2001-2 to 2008-9 regular season on the one hand versus those from the 2009-10 regular season on, there's a very clear trend, one that's disturbing for this year's Jazz. In the earlier years, for teams that had won less than 30 games the prior year, the preseason had a much stronger correlation to winning percentage than prior year's regular season (0.39 vs. 0.06). But in more recent years the prior regular season has a much stronger correlation to winning than the preseason does (0.46 vs. 0.10).

Kind of a bummer.
 
Darkwing Duck said:
How much is that correlation due to influx of a high, ready to play, draft pick?

Jazz don't have that.

What's the correlation to coaching changes?
Click to expand...

Rookies almost never have massive effect on the win total of their teams. I posted about a nice ESPN article about it a while ago that summarized it pretty well:
https://jazzfanz.com/showthread.php?32535

The "NBA ready" rookies who help you "win now" are to a huge extent a myth. It usually takes 2-3 seasons for the rookies to start becoming serious contributors.
 
Stifle Tower said:
So predicting more or less than 23.5 wins for this team is truly a gamble?
Click to expand...

Not necessarily. Based on after-2009 data, the average win total of all teams winning less than 30 games is 23.7. The following year the average wins for those teams is 30.9. Use your discretion to decide whether any particular less-than-30-win team will be above or below the 7-win gain average for the following year.
 
Stifle Tower said:
So predicting more or less than 23.5 wins for this team is truly a gamble?
Click to expand...

Starting with the 2001-2 regular season, 74 teams have won less than 30 games (or lockout equivalent). Their average win total the following year is 8.84 games higher. This breaks down as follows:

30+ more wins, 2 teams
20-29 more wins, 10 teams
15-19 more wins, 10 teams
10-14 more wins, 9 teams
5-9 more wins, 14 teams
0-4 more wins, 18 teams
1-5 more losses, 6 teams
6-10 more losses, 3 teams
more than 10 more losses, 2 teams

If you do the same thing starting with the 2009-regular season, you get:
29 total teams
7.16 average wins more the following year
Breakdown:
30+ more wins, 0 teams
20-29 more wins, 3 teams
15-19 more wins, 2 teams
10-14 more wins, 5 teams
5-9 more wins, 4 teams
0-4 more wins, 12 teams
1-5 more losses, 3 teams
6-10 more losses, 0 teams
more than 10 more losses, 0 teams
 
Wins and losses during the preseason was never a predictor of our regular season record when the Jazz had Karl and John, so why should we think it has relevance now? However, I do believe that the changes Quin implemented for the preseason -- as shown by our improved offense and defense -- will have a positive impact on our regular season record. Will we win more than 40 games? Probably not, but 40 wins doesn't seem so far fetched now.
 
I'm almost certain I've underestimated what this team can do. They look half decent. Then again, last pre-season got my hopes up for Kanter and then he took a dump on them.
 
Ellipse said:
Wins and losses during the preseason was never a predictor of our regular season record when the Jazz had Karl and John, so why should we think it has relevance now?
Click to expand...

Veteran teams approach the pre-season much differently than a team like the present day Jazz. On vet teams veterans use the pre-season to get themselves into game shape (although Malone was probably in game shape 365 days a year) while 6 or 7 guys fight for 1 or 2 roster spots. W's & L's mean little to nothing because you're dealing with a known quantity.

Now we have a young team and a new coach who is essentially a doctor injecting a new serum into the franchise (to counter act Tyrone Corbin's full blown AIDs-like affliction on the team). And the reaction has been mostly positive. It could very well mean nothing and it could mean something - right now I'm leaning towards something.
 
You must log in or register to reply here.

Share this page

Top