Phoenix Suns Best Defense, Detailed Analysis Shows Improvement

Like the awful movie starring Eddie Murphy, the Phoenix Suns have only been able to offer up mediocrity at best, and gut-wrenching displays at worst, on the defensive end of the floor this season. Recent grumblings hit new heights, as calls for firings, benchings and even tomato-ings. A seven-game skid hit rock bottom in a loss to a very poor Orlando team by nine points. All hope was lost and blame needed to be assigned.

Yours truly even wrote that the Suns flat out suck, after an arduous analysis of every possession played over the skid. Check out that article for some interesting insight as to why I came to that conclusion.

Then something interesting happened – we won. And against a team we lost to during the streak no less. AND against a team that has not only a winning record, but is one of the top teams in the league this season.

Hope restored.

Following that big win, everyone proclaimed that heads could remain firmly intact. It was surmised that changes were made that clearly worked, and that our performance was better than adequate. All of a sudden, our defense was not horrible, but decent. Our lineup and rotation issues have been solved. We can now all rest easy knowing that we no longer will drop to the bottom of the cellar, nor rise to the level of a playoff team and that our draft pick will once again be a middling one [wait, nobody wants that].

Not so fast!

Careful analysis of both games show some very interesting things, and lead me to believe that while some areas were improved over the first game, the Suns were helped considerably by Memphis in Game 2.

Keep in mind, the first game went to overtime and was back and forth throughout. That in and of itself was a good thing, considering a 40-point shellacking against a 7-17 Detroit Pistons a week prior.

In Game 1 against Memphis, we actually competed, and against a very good and balanced team. Memphis, might I remind you, has two quality bigs [one who is probably the best low post scorer in the game] a quality athletic point guard and an all-star wing that can clearly score on anyone [wouldn’t you love to have Gay].

Obviously, one conclusion I have drawn from this small sampling of games against Memphis is that by some weird doing, we actually match up well enough to cause Memphis problems. How else can you explain why both games were within one possession of going either way?

So for fans, the fact is, we played well against them the first time, despite the loss. In fact, if you look at that streak of losses, then Suns played fairly close to NY, Dallas and Orlando. Were it not for some stretches of the game where we mentally checked out, some or all of those could have gone the other way, and rather than an 8-15 record, we might be 15-8 and talking playoff seeding [and laughing even harder at LA]. Keep in mind, we have lost eight games by 5 or less points or in OT.

So instead of getting all "verklempt", maybe we should "simmer dahn nah" and keep some perspective.

Back to the game, since I did spend over 16 hours compiling all of this data…

First, let me give you a brief review of some findings from the first game, then I will go over what changed to the second. Remember, that I was focused solely on the defensive end, so everything I am talking about is Suns defensive possessions [or Memphis offensive possessions]. That is not to say I looked at things in a vacuum – I accounted for situations where our offense might have affected the defense or vice versa.

Note: This was written pre-Jazz game…

A NOTE ON METHODOLOGY:

As some of you are more detailed oriented [and others more skeptical] I feel it is necessary to provide some level of detail regarding how I ultimately came to some of the conclusions expressed herein. Therefore, to spare you time on the front end with all of this boring detail, I will place such information at the end of the article for your reference. I have not included my excel spreadsheet. If you would like to see that, write me a big fat check.

GAME 1 NOTES:

Here is the period-by-period average of our overall defensive rating for Game 1 as described in the methodology at the end of the article. Remember, I rated every defensive possession from 1 [worst] to 5 [best], so 3 should be average, or decent enough.

As you can see, our best performance defensively, in terms of how we actually played [but not the resulting outcome] was during the second period [27-27 actual score] despite the fact that the first quarter [25-18] actually turned out better for us on the scoreboard.

Keep in mind the fact that when you are dealing with human beings, you cannot predict any given outcome for a small sample size. What I mean is, PJ Tucker played incredible defense on Rudy Gay, yet Gay still scored, and conversely I saw Gortat play horrible defense on Gasol, only to see Gasol fumble the ball out [no credit to Gortat] and end the possession with a turn over. Great defense doesn’t always result in great results, and vice versa.

Another interesting thing to point out here is the evidence that we start games and the second half poorly, and this is possibly due to our lack of defense – it certainly shows in the numbers.

Maybe we should be focusing on that with regards to our lineups, rather than how well the offense is clicking.

The numbers here show suggest the reason we lost this game was not adequately defending in OT. One might think that the fourth quarter, where we gave up a four point lead, was the cause, but we actually played better defensively in that quarter. I am suggesting that Memphis simply executed better offensively against our improved defense in that situation.

Remarkably enough, during the third quarter, the Suns blew seven possessions with turnovers, which attributed to some of the decline in defensive performance [as their transition defense clearly was bad]. However, despite all of those turnovers, there wasn’t a significant change in actual score, as we still held a four point lead going into the fourth quarter [as opposed to a seven point lead].

So now let’s quickly look at Game 1 and how all of the possession’s played out:

As you can see from the chart above, Memphis’ offense was fairly evenly distributed among types of plays to initiate their outcomes. One thing to note was that the combination of PNR, ISO and PEN showed that Memphis got in the lane regularly and attacked the rim aggressively [keep that in mind].

The Grizzlies were adept at converting Post and Isolation plays, with those plays resulting in a score more than 60% of the time. You will also notice that the Suns defense against those plays was not stellar, falling significantly below average [3.0 is average] with 2.55 and 2.37 respectively.

The Suns were able to get closer to adequate on guarding dribble penetration [2.90 rating], which might surprise those who claim our guards get scorched by quicker more talented guards. Conley is very quick and Gay can do what he wants on the floor, but this shows we did a decent job [or just below decent].

Additionally, the concept that our big players might be the issue rather than the guards, is evidenced by the 2.71 rating on our pick and roll defense [versus the better rating on dribble penetration, showing that the guards are doing a slightly better job guarding their man when they are not being picked off].

Getting more detailed on the PNR, I decided to test my earlier assumptions that the Suns are far better off hedging or trapping hard on every pick and roll. Obviously as a coach, I intuitively believe this as a general rule [and not in exclusion of making adjustments for certain teams/players that are extra capable of beating those traps], but I wanted to see if the results match.

Well, as they say in Mythbusters – CONFIRMED!

Of the 21 PNR plays, the Suns hedged/trapped hard on 9 plays [and sagged on 12]. When they did, Memphis converted a score on only 3 of those plays [33% conversion] and allowed a total of 8 points.

Conversely when they sagged, Memphis was able to convert a significantly greater amount of possessions [7 or 58%] and allowed a total of 14 points. On a point per possession comparison, you are talking a difference between 0.89 on hedges and 1.17 on sags, a tremendous difference. Even more alarming, the rating of how well we performed defensively between the two was night and day.

Our rating on hedge possessions was a whopping 4.11, which is close to dominant, while our sag defense rated a paltry and pathetic 1.58. Why would we ever use sagging as a strategy based on those numbers? We clearly are not good at playing that way, and the results from hedging show we are both great, and successful at producing positive outcomes.

Finally, another observation I had is that our offense must not be that bad. We stayed in the game with a top team all while playing below average defense. Our offense had to have taken up the slack to be able to get it tied in regulation. In fact, the Suns led most of the game [other than in the beginning]. All of this talk about scoring, the need for a go-to player, touches, etc., and it occurs to me that it really might just be the defensive end that needs work.

THE DIFFERENCE IN GAME 2

The first thing you might notice from the chart above is that our overall defense improved from what I call below average to just about average. While you might not get overly excited about, it is the detail you can smile over.

First, while we continued to start the game without much defensive prowess, we seemed to correct our third quarter lull by posting a hearty 3.4 rating in Game 2’s third quarter. That clearly is above average, and while not great, is good. You will also notice that our entire second half was solid. But surprisingly enough, we still only managed to stay close with Memphis and eek one out.

Looking at the breakdown of possessions yields a treasure trove of interesting tidbits. First is the incredible disparity amongst possession types. As opposed to Game 1, Memphis utilized the PNR almost twice as much, and at the expense of evenly distributing how the instigated their opportunities. If you remember, in Game 1 they attacked the rim aggressively through PNR, ISO and PEN. I say attacked because Memphis was able to get in the lane on PNR by both our sagging defense [allowing them into the lane] and their aggressiveness in trying to get there. Conversely, in Game 2, they seemed to lack aggression going to the rim, choosing not to penetrate or ISO as much.

The other area that I found intriguing was the fact that Memphis only had seven, SEVEN post plays in Game 2 [Game 1 they had 20 – and converted 60% of them]. You might say our dramatic improvement in our defense [3.57 rating in game 2 from a Game 1 rating of 2.55 – nice!] contributed to their choice to ignore throwing it to the block, yet they still converted about the same percentage of post plays they ran. This is why I believe that despite our improvements on the defensive end, Memphis did us a favor by going away from their clear advantage over us.

The fact is, that when you look at the conversion rate for PNR, POST, PEN and ISO, there isn’t a significant change. This means they are good at, and we are not good at, those types of plays, no matter how much better or worse our defense might have been from game to game. This shows you that matchups in this league are very meaningful. Something about the Suns and how we are constructed gives Memphis a difficult time on the PNR. In Game 2, they mistakenly went to that as their bread and butter, when they should have focused on throwing into the post and isolating Rudy Gay.

In Game 2 you should notice a trend – that we are not good in transition defense.

In both games we were pretty awful, and that isn’t because we get scored on. We are awful because too many times, Memphis pushed past us for no real reason other than our laziness or arguing with the refs. While that happens with a lot of teams at times, we consistently played the break poorly, making classic errors like not stopping the ball or simply losing someone behind us.

It doesn’t help that turnovers and bad browns, I mean shots, contribute to a disadvantage in transition. Yet the ratings aren’t low because we didn’t have numbers, but for the reasons above. It is expected that we will give up fouls or points on the break. What is not expected is the number of times you groan watching them out hustle us. The Suns need to fix this, because there is no excuse for it.

Another tidbit is this idea that our newfound zone defense. It turns out, at least against Memphis, that our defensive rating of 3.11 was solid in comparison to anything we did in Game 1. The fact that Memphis only converted 22% of possessions when we went zone, and scored only 4 points on 9 possessions, showed that Memphis has a significant issue [if this small sample actually represents how they play against the zone].

I think I might throw some more zone the next time we play them and see if that holds. Certainly, this lends credence to my theory that Memphis was not attacking like they were in Game 1, as they settled for a lot of perimeter play in Game 2 – helping the Suns.

Finally, after my scathing diatribe about how the Suns need to improve their pick and roll defense, let’s look at whether they were listening to me (because we all know they were).

What is interesting here is that despite a significant increase in the number of PNR possessions, Memphis still managed to convert slightly more of them in Game 2. One might conclude then that their PNR defense did not improve. I would contend that Memphis actually improved their execution of the PNR in Game 2 rather than stating that the Suns did not improve their defense. The evidence shows this.

First, the Suns hedged on a significantly greater proportion of possessions [51% in Game 2 to 43% in Game 1]. In fact, the grizzlies converted FEWER possessions when hedged than the first game. Again, it is abundantly clear that hedging/rotating is key. Despite more than double the amount of hedged possessions, the Suns managed to hold Memphis to only 2 more points off those plays [Game 1 PPP 0.89 to game 2 ppp 0.53]. If the Suns could manage to hold teams to even 0.75 ppp, they would be the single greatest defensive team in the history of mankind [don’t hold me to that factpinion].

The real reason the Suns were able to play better defense [rating increase] yet still not decrease the Grizzlies conversion rate were the possessions they sagged on the PNR. The suns gave up double the points, despite an increase in sag possessions of only 50%. In other words, they allowed 78% of saggy possessions to score in Game 2 versus 58% in game 1.

Clearly, I am not going to get off my soapbox about this hedging thing.

The disparity between how well we defend [in ratings and in results] is so wide, you have to be bit nuts not to take note. While some would argue the “well, it depends on the players you are guarding” chestnut, I believe that would only matter in a situation where the rating and result disparity were closer and there was more risk involved with making that type of strategy choice. When the numbers are so lopsided, you are really able to risk the occasional bad play or “beat” here and there knowing that on the whole, the odds are overwhelming in your favor to do it.

One final anticlimactic and completely obvious observation [from looking at all of these possessions]: Jared Dudley is a freakin’ stud. He made so many plays in the 4th quarter in Game 2 that I curse any of you expressing your willingness to trade him. Celo said it best – “forget you”.

____________________________________________

METHODOLOGY:

The first thing you have to understand is that I do not have access to Synergy Sports Tech, or any sort of wonderful technology that allows me to index video automatically and dump that data into any database. I also am not sure that what that wonderful technology does actually can replace what a person, who is knowledgeable in regards to the game, can do. In fact, as I did this, I realize how subjective any action on the court can be interpreted, and that someone with less than stellar experience could be swayed to view certain actions in a way that does not actually fit with a fundamental laws of basketball. Yet I digress. Bottom line is that I watch games much differently than someone who simply enjoys basketball or plays it. It is in my nature.

Using my DVR, I reviewed the game - every second of it. In fact, I rewound and re-watched every play a couple of times to make sure I didn't miss anything. Prior to viewing, I put a lot of though about the plethora and variety of items I would find during a possession and created a chart to fill out. Obviously, if I had infinite amount of time and money, I could have broken out many more things, such as how long any individual guarded another or the ball, who was on the floor and when, and every single minute detail. Rather than do that, may main focus was to capture the main essence of the defensive play.

For example, while a pick and roll [PNR] might have been employed on a possession, unless that PNR was the instigator or initiator of the coming result [score or change in possession], it was not as relevant to the cause of the end result of the possession. So if Memphis ran a PNR that was poorly executed, or even guarded well, but then decided change course and throw the ball to the post and isolate Randolph for 8 seconds of the shot clock for a score, the main offense used to initiate that score really was a post play, and the PNR was not factored very much into the play [although if they played it really tough, I did note that and would factor that partially into rating the overall possession a little higher].

For each defensive possession, I notated the type of play used to initiate the main action. I also notated the main defensive players involved, what they did, both right and wrong, as well as my “opinion” with regards to how the defense executed [my opinion was based on my hundreds of thousands of possessions I have coached over the years – and my ability to understand fundamentally why things occur on the floor]. Additionally I noted the results of the play, including special note of how the PNR was played, whether we hedged/trapped or sagged.

I tried to account for all situations within my notes – such as times when great D only resulted in a lucky score, or conversely bad defense ended with a steal or a miss, and everything in between. I then rated those actions on a scale of 1-5, with 1 being terrible defense and 5 being great defense. Any great defense that still resulted in a score got a 4 rating. Any stop with terrible defense usually earned a 2 rating [except for possessions, for example, where we clearly played extremely bad defense only to have Memphis throw the ball stupidly out of bound, but not of our doing – those got a 1 rating].

I then took all of those notes, entered them into Excel and created sortable tables. Once I had all of the data in the tables, I was able to organize them and create some analysis which you see above. Essentially I recorded how many defensive possessions, exactly what plays were used and the results, and rated every possession. Then I worked some math and was able to come up with some metrics to use for comparison among both games played against Memphis.

For the charts, here are what the headers mean if you need further explanation:

TYPE = action that initiated the result on the possession

PNR = Pick and Roll

POST = Post play

ISO = Isolation or a play in which the player created on their own [not in the post]

PEN = Any penetrating move not using a PNR, but off of some sort of team movement or set [not isolation]

SW = possessions where no clear play was used, no picks or curls, just perimeter ball movement to open player

CURL = Really any play from a curl or a set of picks off the ball

FB = Fast Break offense

OTHER = Things like offensive rebounds, offense created from a turnover, or mad scrambles that don’t fit with any other category

# = number of times that occurred

% = Distribution of the offense – what percentage of the offense came from that type of action

PTS OFF = Point generated from that type of action

CONV = How many of those plays led to a score

CONV % = Rate of converting those plays for scores

DRTG = The average rating of those plays based on the rating system above