Should the Celtics call a players-only meeting?
Jorge Sierra of HoopsHype wrote an interesting post today about the effect of players-only meetings on team's record. While he recognizes the two basic issues with this analysis (i.e. limited sample & correlation does not mean causality), it's interesting that on average such meetings have produced positive results. To be specific, in 10 of the cases it yielded positive results, in 1 the short-term benefit was cancelled during the rest of the season, and in 4 things took a turn for the worse. I believe that such meetings actually do work, and further statistical analysis shows that it produces significant positive results at least in the short run.
Let's first hypothesize why such a meeting might work. There are several obvious reasons: It emphasizes the urgency of the situation, players possibly feel freer to talk their minds, it offers a "new page" which usually works out fine (think coaching changes etc.), there is an extensive focus on the problems... Basically, it is not bizarre to think that any correlation would be meaningless.
Now, there are several problems with the data presented here: There can be many factors such as difficulty of schedule, home-away games, roster/personnel changes, injuries... You name it. However, assuming ceteris paribus exists, we can conduct a paired T-test and test the null hypothesis that the treatment, i.e. players-only meeting doesn't work at all. [Note: That something is statistically significant does not mean that it is true, it means that you cannot reject it as being untrue. The nuance is a bit tricky.]Results are as follows: The games before the meeting and the first 10 after are significantly different, so one can confidently say that such meetings do have an effect in the short run. However, the results are not significant in the long run, i.e. the rest of the season. Obviously, a study that takes into account multiple factors such as ones listed above would explain this phenomenon better, but I'll take this.
So to answer the titular question: Celtics could use a players-only meeting and it will probably be beneficial in the short run. Why? We have recently read that there are some trust issues between the players, the meeting last year seemed to work quite well as it fueled the Celtics comeback in the second half, Garnett said that this is the time that they'll know who's with them and who's not etc.
In a nutshell, there are signs that the Celtics might be needing such a meeting. However, you cannot have multiple such meetings because that will reduce its effectiveness as one can imagine. It's important to time it right.
I say that the Celtics wait to see whether Bradley's return shakes things up a little. If not, then it will surely be time to talk things out.
Let's first hypothesize why such a meeting might work. There are several obvious reasons: It emphasizes the urgency of the situation, players possibly feel freer to talk their minds, it offers a "new page" which usually works out fine (think coaching changes etc.), there is an extensive focus on the problems... Basically, it is not bizarre to think that any correlation would be meaningless.
Now, there are several problems with the data presented here: There can be many factors such as difficulty of schedule, home-away games, roster/personnel changes, injuries... You name it. However, assuming ceteris paribus exists, we can conduct a paired T-test and test the null hypothesis that the treatment, i.e. players-only meeting doesn't work at all. [Note: That something is statistically significant does not mean that it is true, it means that you cannot reject it as being untrue. The nuance is a bit tricky.]Results are as follows: The games before the meeting and the first 10 after are significantly different, so one can confidently say that such meetings do have an effect in the short run. However, the results are not significant in the long run, i.e. the rest of the season. Obviously, a study that takes into account multiple factors such as ones listed above would explain this phenomenon better, but I'll take this.So to answer the titular question: Celtics could use a players-only meeting and it will probably be beneficial in the short run. Why? We have recently read that there are some trust issues between the players, the meeting last year seemed to work quite well as it fueled the Celtics comeback in the second half, Garnett said that this is the time that they'll know who's with them and who's not etc.
In a nutshell, there are signs that the Celtics might be needing such a meeting. However, you cannot have multiple such meetings because that will reduce its effectiveness as one can imagine. It's important to time it right.
I say that the Celtics wait to see whether Bradley's return shakes things up a little. If not, then it will surely be time to talk things out.
