Why we lost - Part II: Defense and rebounding

Why we lost is a series of posts that looks at statistical data and tries to analyze what the beyond-the-obvious factors for the Celtics' below average regular season performance were. The main goal is not to pinpoint what went wrong but to provide insight for potential improvements and material for always-fun "what if?" games.

Wow. I posted the first part of this series a month ago, and a lot, I mean a lot has changed since then. The bulk of my individual player analyses would have focused on Garnett, Pierce and Terry who are all gone. I will try to make these posts less nostalgic and more optimistic as much as I can, a task that might be quite impossible while analyzing Garnett, alas, but for today my main focus will be on defensive and rebounding stats as a team.

I. Rebounding

The Celtics and rebounding has always been one of my favorite topics to discuss, and I'm quite joyful to have an 82-game sample to play with and reach conclusions from. As I have previously written, I believe that even though the offensive rebounds are important when they turn into second chance points or steal the opponent's momentum, they don't categorically translate into wins or losses if you don't care about them by design, which was the case for the Celtics and will probably continue to be so under Stevens' coaching (see clause 3)

Anyway, let me cut to the chase: The Celtics grabbed 7.8 oreb/game in their wins, and 8.4 in their losses last season, which shows that grabbing offensive rebounds is not a factor for these Celtics. Defensive rebounds, on the other hand tell a different story: When the Celtics won, they grabbed an average of 32.7 boards/game compared to 29.8 in losses. There are two possible explanations for this: a) The Celtics gave up more offensive boards and b) the opponents made more of their shots. Well, apparently the latter is true. The Celtics' opponents grabbed 12.3 offensive rebounds on average when we won but only 10.5 when we lost. However, their FG% went up from 41.7% to 46.6% under the same scenario. OppDREB/game is also higher in losses than in wins, but that's also due to the fact that the Celtics miss more shots in losses than in wins.

Boxing out is important.
To sum up, yes, the Celtics don't grab as many rebounds as the other teams do, yet there are more important factors than rebounding that determine who wins the games. After all, the two teams that made it to the Finals last season ranked 28th and 29th in OREBs. As long as you take care of rebounds on the defensive end (the Celtics ranked 12th in that category but could have ranked higher) things should be fine.

II. Turnovers

This is an interesting one. It's quite obvious that turnovers decrease a team's chance of winning as long as the other team isn't less careful with the ball. However, if the Celtics were playing at home last season, TO statistics were absolutely meaningless: The Celtics turned the ball over 14.9 times in home wins and 14.8 times in home losses. The opponents? They turned the ball over 16.6 times in home wins and 16.5 times in home losses. I really do not know what to make of these numbers. However, for the road games, turnovers did made a difference: 13.2 in wins, 14.9 in losses for us; 16.0 in wins, 13.1 in losses for the opponents. Yay, common sense is not lost.

A monthly analysis reveals another interesting information: The Celtics turned the ball over 12.8/game in February, a number significantly lower than the averages in other months. The players whose turnover numbers were lower in February compared to their season averages were Terry and Pierce, but it is also fair to say the fire Rondo's absence lit under the team shouldn't be underestimated. If only the team didn't get too used to playing without him...

III. Steals

Always a pleasure to watch.
The number of steals isn't necessarily an indicator of a team's defensive abilities; after all the Pacers and the Bulls rank 22nd and 23rd respectively, yet stealing the ball is supposed to be a forte of the Celtics backcourt. Steals indeed made a difference: 8.9/game in wins, 7.6/game in losses. Playing at home or on the road didn't matter, the Celtics' steal average was always higher in wins.

Of course there has to be a piece of statistics that defies common sense in each category, so here it goes: The Celtics' opponents stole the ball 8.1 times in our road wins, yet 7.6 times in our road losses. That can be explained by the fact that the Celtics' steal numbers were higher in road wins (8.2) and lower in road losses (7.1), yet still it's an interesting piece of information as it shows that steals make more sense in relative terms than absolute ones.

It should also be noted that the Celtics' steal numbers were the highest in December and January, 9.1 and 9.0 respectively. It's no coincidence that the Celtics' steal leaders, Rajon Rondo and Avery Bradley played together in January, yet the increase in December can be attributed to Paul Pierce's impressive performance.(2.0/game)

IV. Blocks and Fouls

Blocking shots is a department the Celtics can definitely use some help as they rank in the bottom 10 with 4.5 blocks/game. It is also quite clear that block margin and wins are directly proportional: The Celtics had a blocking margin of 0.6 in wins vs. -0.8 in losses. Even though this figure is quite close for road wins and home losses, the difference is quite significant in home wins (5.3 vs 4.4) and in road losses (3.9 vs. 5.4) Now, I don't think that blocked shots is a direct determinant for wins or losses but it's mostly important as a psychological factor: crowd motivation, inside protection, penetration deterrence etc. That's why the difference is most significant in home wins and road losses.

If only...
The same reasoning goes for the fouls to a certain extent. On average, the Celtics fouled their opponents more than their opponents fouled them (21.2 PF/game (4th in league) vs. 19.4 PFD/game), yet the net difference was 0.8 in wins vs. 2.7 in losses. The only team with a worse foul margin to make the Playoffs were the Warriors. This number got even more significant for road games: If we won on the road, we fouled the opponent only 0.4 times more on average, yet if we lost, the foul discrepancy was 3.2/game. The number of fouls called against us were also lower in absolute figures in road wins (19.3/game). Basically, we won if the referees ruled the game more leniently: The Celtics loved grit-and-balls games, to quote Kevin Garnett.

One last nugget: The number of fouls called both for and against the Celtics have decreased over the season, which means that either the teams have adjusted or the referees got more lenient.

V. Conclusion

As it was the case in Part I, the results are hardly surprising yet there are a few gems. To summarize:

- Offensive rebounding doesn't matter as long as it is by design, yet it is important to grab the defensive boards, something the Celtics do decently.
- It won't hurt the Celtics to take better care of the ball, especially on the road.
- Shot blocking numbers are quite meaningful for home wins and road losses, and the Celtics are in need of even more inside presence now that KG is gone.
- The referees don't love the Celtics, and it shows especially on the road.

I will focus on individual performances in the next few posts.
Previously on Why we lost:
Part I - Shooting in wins and losses

Follow semioticus on Twitter - @semioticusCL