There has been a lot of back and forth regarding Bradley’s relative worth, who should be starting in the Celtics back-court when Rondo comes back and who’s worth trading or keeping, I wanted to try and take a look at what the Defensive numbers (or the numbers I picked) say about Bradley, Crawford and how they match-up against the rest of the players at their position.
Right out of the shoot, I’m pretty much a neophyte at Basketball Advance stat analysis , but I CAN read (that always helps) I can listen to others and I can usually make the jump to making my own decisions about what things mean after immersing myself in a subject.
I wanted to start with defense. Why"? Because we’re always hearing anecdotal data about what a superb, game changing defensive player Bradley is. He was named to the **"all defensive" second team last year, Jameer Nelson got some headlines about asking him to "back off" on his "D" during a game and there are a few other apocryphal stories regarding how good his defense is. These stories and awards are all pretty subjective but is his reputation as a defensive wiz earned?
**The voting panel consisted of the NBA’s 30 head coaches, who were asked to select NBA All-Defensive First and Second Teams by position. Coaches were not permitted to vote for players from their own team. Two points were awarded for a First Team vote and one point was awarded for a Second Team vote
So what do the numbers say? My first challenge is to figure out what to look at. Here’s the criteria I used (actually stole from an article I found) and why
Defensive Rating (DRtg) – From Dean Oliver’s book Basketball on Paper. Defensive rating simply shows how many points a player allows per 100 possessions. It eliminates pace and minutes played so it should be fair (if you trust Dean Oliver).There is some supposedly fuzzy math in his formula & there’s a GIGO factor involved in the raw numbers but it’s been accepted as a useful tool in measuring defensive efficiency….the lower the number, the better.
Defensive Win Shares (DWS) – Defensive Win Shares takes all facets of the defensive game into account and is tempo-free. It’s reported downfall is that, while it is supposedly pace and tempo free it does reward players for more minutes played.
Ok, so you may be saying to yourself, laa-dee-daa or "Dude, those are entirely the wrong things to look at". to which I say, "do your own analysis then or eat some more Cheetos"
So now what. Well Basketball reference allows us to do some pretty cool filtering when it comes to stat analysis so I can take those stats listed and compare how a Bradley or Crawford compare to each other and/or where they rank within the league.
So what’s a fair comparison? I decided to look at players who were identified strictly as guards, who averaged over 20 minutes a game (that should capture all starters and 3rd & fourth guards without letting some wild cards sneak in) and those that have played over 300 minutes so far this season (why? In case some really good guy like a Derrick Rose who got injured and only played a few games went missing…and the 20 minutes a game should eliminate any "frauds")
What do I find?
Crawford is at #18 with 0.8 DWS (he’s basically @ # 15 because 3 other guys are at 0.8 & 3 other guys are @ 0.7…the database appears to rank them alphabetically when they are tied…..whatever)
Well, that doesn't match-up with what I thought I’d see….so let’s move on, maybe it’s just a fluke. Let’s go to Defensive rating.
Here using the same 83 guys we find Bradley at #25 with a 105 rating. For comparison the "best" rating is Manu Ginobili at 98 & the worst? Again, the lovely and talented Joe Johnson at 115.
Where’s Crawford? At # 31 with a 106 rating once again falling victim to alphabetizing, he might as well rank # 26.
If you go back to last year, Crawford & Bradley are basically still neck & neck using those same stats. Rondo beats them both, hands down. Weird stuff
So what does this all prove? Maybe nothing, there are lies, damn lies and there are statistics, but it does allow us for some base to reject the notion that Bradley’s defense is "game changing" or so over the top better than other players that his offensive showing can be ignored or that an average rating offensively is ok because of his crazy tough D. (which is another conversation, let’s just say that Hollinger’s PERformula puts him at 12.04 where average is considered 15, Crawford is at 19.6) The caveat to that is that Bradley’s has been much much better offensively since being relieved of his PG duties
You can look at that yourself at http://insider.espn.go.com/nba/hollinger/statistics
Just another mention of PER some hate it, it doesn't take defense into account much but We accounted for that elsewhere.
Fell free to rip this apart or ignore it, I found it useful because I feel I can actually gauge How good is good even with the relatively small sample size so far this year 25% of the way in.
Here are the links to the analysis plus an additional one. Steal Percentage that was mentioned in a reply: