Strangely enough, Bradley's offensive units were more efficient than Allen's last season.

We are fully entrenched in the NBA’s dog days of summer, and for websites that cover solely one team, even such quality websites as Celtics Town, occasionally the same debates get rehashed over and over. So, naturally, I’m going to go ahead and continue to rehash an already rehashed debate; a debate so rehashed, it has become nearly as redundant as my use of the word “rehash” in this paragraph.

For quite some time, I’ve maintained a somewhat controversial position: that the Celtics will not only be able to survive offensively without Ray Allen, probably the greatest three point shooter of all time, but that they will be better without him. A closer examination of Boston’s five-man floor units, from the ever reliable 82games.com, presents compelling evidence in my favor (rephrased: I TOLD YOU SO HATERZZ).

Celtics fans seem to universally agree that last season, a lineup of Rondo/Pierce/Bass/Garnett +Bradley was the best defensive unit, while the same unit replacing Bradley with Allen was the ideal offensive group. Those lineups were the two most popular for Doc Rivers. Combined, they made up 545.8 minutes, as the Allen version played 326.8 minutes to Bradley’s 219.0. Both minute totals make up sufficient sample sizes to draw conclusions based on the results (as opposed to the Bradley/Dooling/Pavlovic/Garnett/Stiemsma lineup that averaged a stunning 1.32 points per possession in a meager 31.9 total minutes this season).

Having established all of that, here’s where things get a little weird: the standard Bradley lineup was MUCH better offensively than the standard Allen lineup. Here are the numbers:

While Allen’s lineup averaged 1.06 points per possession, Bradley’s lineup was .07 higher, at 1.13. In an entertaining (but unrelated) coincidence, the difference between Bradley and Allen’s lineups defensively has exactly the same differential of .07, as Bradley’s lineup topped Ray’s, .94 to 1.01.

Defensively, we shouldn’t be very surprised by these results. Bradley is a better individual defender than even a healthy Allen, and the Allen we saw in the late regular season/playoffs was far from healthy. But the offensive differential doesn’t line up with the standard narrative. It also, quite honestly, doesn’t match up to standard logic. Despite Bradley’s obvious improvements offensively and Allen’s notable aging, we would have expected the better offensive lineup to be the one with the better offensive player. Allen was clearly still better offensively than Bradley. Ray had a higher PER, a better eFG%, and a lower turnover percentage (turnovers per 100 possessions).

So what do these numbers mean? Is it possible that Boston’s offense simply didn’t need to play Allen to be efficient? Actually…yes. Yes it is.

While having an incredible three point shooter like Allen is undeniably a good thing, Allen’s slipping athleticism has made him into more and more of a three point specialist over the course of the last few years. Allen achieved his highest PER in 2003, posting a 23.3. Last season, his PER dipped all the way to 14.8, his lowest since his rookie season. Part of this can be attributed to the ratio of three pointers to two pointers Allen took. 39.8% of his shots in 2003 were three pointers. In 2012, 47.6% of his shots were from behind the arc. And though Allen hit a better percentage of his 3 pointers in 2012 (.453, the highest percentage of his career), the fact remains that three pointers aren’t the most efficient shot. While Allen is a better 3 point shooter than Bradley or Pierce, he isn’t ENOUGH better to make his lineups more efficient than Bradley’s.

One argument that could be logically made is that while Bradley’s lineup may be more effective offensively, Bradley can’t (and shouldn’t) play all 48 minutes. At some point, a player needs to be brought in as a replacement. But when the Celtics tried another less efficient player, Mickael Pietrus, instead of Allen in the same Rondo/Pierce/Bass/Garnett lineup, the unit was extremely efficient at 1.43, though admittedly the sample size was by no means reliable; this lineup recorded just 34.7 minutes together. But in this unit, Pietrus was clearly not the offensive focal point, much like Bradley in his. Rather, he was a complementary piece, while Pierce, Rondo and (perhaps most notably) Garnett took charge. It should be noted that no unit that featured Garnett at center finished with an offensive efficiency of less than 1.06.

By no means am I suggesting that having Ray as a three point specialist would be a bad thing. But I am suggesting that perhaps Allen’s presence represented an offensive focal point that prevented Boston from achieving their peak efficiency. Ray needed his touches, and giving him those touches took higher percentage shots away from other players. Perhaps merely having Rondo, Pierce and Garnett with various other talented players like Bass (and, we hope, Jeff Green, Courtney Lee, and Jason Terry) is enough to make a team excellent offensively.

Feel free to rehash my thoughts in the comments.

Follow Tom on Twitter @Tom_NBA