clock menu more-arrow no yes mobile

Filed under:

My Approach To Understanding Brandon Jennings (And Everyone Else)

Just a few days ago I wrote this article suggesting Brandon Jennings should not be considered a core player if the Milwaukee Bucks want to build a championship caliber team (but why else do teams bother building cores anyways, right?). It didn't exactly go over well. Most people like the youth, potential, competitive fire and everything else that comes with Jennings, and aren't ready to consider the possibility that he might not be the proper player on which to pin the future hopes of the franchise. I get that. Honestly. Now when it comes to the point that when one other Bucks fan on the planet actually shares my skepticism (victor s) it is so inconceivable that a wacky alternative explanation -- one of us created a fake internet personality and posted similar opinions under both screen names to create a small two-person agreement -- is explored, that confuses me. Anyways, before I become the guy for whom everyone wants to take my temperature on Jennings after every play and every game, I might try to explain my general approach and how the article came into the world.

A great place to start is the closing paragraph of the Jennings article:

So here we are, 2600 words later but not far from where we started. Once again, the tantalizing results are contained in a small sample size, while the overwhelming weight of the evidence casts a dark and expansive shadow on Jennings as a player. It's almost odd that I timed this post for the best stretch of play Jennings has ever shown, but maybe he can't do what I think an offensive centerpiece or core player needs to be able to anyways. With Jennings it feels like death by a thousand cuts. Never anything major to signal crippling inefficiency or limits on his future development, but rather a creeping mediocrity that largely goes unnoticed amid a sea of other more expensive and eye-catching wreckage.

If you want to know how I feel when Brandon Jennings goes off for 36 points on 15-26 shooting, with 5 assists, 2 steals and 3 turnovers at Madison Square Garden in a rout of the New York Knicks, it's exactly like you feel. Ecstatic. Amazed. Glad to have witnessed it. More than happy to write about it. I love when the Bucks win. Each game is its own separate experience to me, because almost anything could happen in a single game. The unpredictability is a big part of the reason why we all love sports.

Each game has a unique story about why a team won or lost, and that's the allure of our game recaps. We relive the experience and help people understand what exactly happened on a given night that contributed to the win or the loss. The focus is not on what a player has done over his career, but how he showed up that night. We still bring in our own understandings of each players true talent level to contextualize and rationalize the event, but the event is what matters. If John Salmons had scored 46 points on 21-25 shooting when the Sacramento Kings played the Bucks back on January 5th, would you have changed your general opinion on him as a player or his true talent level? Of course not. Anything can happen on any given night.

Then again, the Bucks' start to the season isn't exactly unpredictable. Neither is the start by the Miami Heat or the Chicago Bulls. Why? Because the top players on the Bulls and Heat have established resumes as elite players in the NBA. They have proven over time they are a cut above the rest, and while anything could happen to them in a single game, they will beat opponents more often than not when given enough chances. This broad view is what colors my long-term opinion of Brandon Jennings and every other player in the NBA. Allow me to explain it in different terms.

Let's say you had a quarter you knew to be perfectly weighted, such that the probability of it landing on heads or tails is 50-50. However, with each flip their are two mutually exclusive options. The coin can either land on heads or it can land on tails. It can't land on both at the same time. In other words, you can't flip and achieve the mean on any single event. These are random binomial outcomes with normal distribution: every event is random (100% or 0%), but still centered around a true mean (50-50). Each toss is a variation based on luck.

Probabilities do not tell you what is going to happen, but they do tell you what is likely to happen. If you toss a coin 20 time and get 15 heads and 5 tails the outcomes was unlikely, but the probabilities have not changed. Over time you would expect the ratio to regress to the mean (50-50). The coin is still not more likely to be tails on the next flip. Rather, the odds stay exactly the same, and in truth any sequence of twenty tosses is just as unlikely as the one described above, even if the distribution looks random.

What the heck does any of this have to do with sports or NBA basketball? Plenty. Players aren't coins though, and we can't ever precisely know their true talent level (ex: Brandon Jennings' make-miss probabilities), but that's a wonderful thing. We can never have enough information to truly know a player's exact talent level (even the coin probabilities are premised on an infinite number of events), but with each game we get insight into a player's revealed talent level. These are the stats we all look to, and this is why we look at them. At some point, there is enough information to make a reasonable estimate of a player's talents going forward.

Just like with the coin flip, no NBA player can shoot their career percentage on a single shot. Only two things can happen: they either make the shot (100%), or they miss the shot (0%). Yet no player shoots 100 percent, and no player shoots 0 percent. the random individual occurrences are center around a true mean that reveals its self more concretely with each passing shot or outcome. In baseball, the minimum threshold for making a reasonable estimate of talent level is around 700 plate appearances. I just happen to think the same minimum threshold in NBA basketball might be somewhere around 2300 shots.

Jennings has been given more opportunities to reveal his talent level than nearly every other player during his career in the NBA with regard to shots, so it seems reasonable to me to use this large sample size as the basis for projecting his true talent level with a decent amount of confidence. What I mean is that I don't expect Jennings to shoot spectacularly better (ie- to a level well above-average for his position) over his next 2300 shot attempts compared to his first 2300 attempts. I haven't looked closely into this 2300 shot threshold yet, but I honestly can't think of any player who made huge jumps from league-worst to above-average, and certainly not one who fits the physical profile of Jennings.

Still, he is young and players are not coins. The mean is bound to change over time, because unlike the coin we can never know the true talent level of a player, only the revealed talent that emerges over time. The odds change based on situation as well, so better decision making with the ball we have seen this year might actually help to improve his established talent level over time if he can keep up the shooting numbers. Brandon Jennings is never going to shoot 38.5 percent on a single shot, but when you give him a couple thousand shots, he did. My problem is that common sense still says he is due for a regression over the course of the season, at least if you believe 2300 NBA shot attempts tell you something meaningful about a player's talent level. I do. It's as simple as that.

I don't dislike Brandon Jennings. I dislike the Bucks taking the long odds in the gamble and hoping he is some sort of outlier in the NBA landscape that can turn 2300 shots worth of league-worst type performance and suddenly turn around to produce all-star shooting efficiency in his next 2300 shots. It's a dangerous gamble for a franchise teetering on a rebuild to give big money on Jennings and pray for him to be that once-in-a-decade outlier. That was the point of the article. Trusting the odds doesn't always pay off, and I could have lost a lot of money to someone whom I bet heads would not come up 15 times in 20 tosses of a coin, by my belief is that playing the odds is still always the right thing to do and will pay off over time. To me, it's what an organization serious about building a championship contender does.

I would love to be wrong about this. I will be the first one to happily admit the mistake when Jennings is leading the Bucks to deep playoffs runs and piling up all-star appearances. I desperately want the Bucks to become a great team during my lifetime (born in 1985), and if Brandon Jennings can overcome the pile of evidence that says he can't be an impact player I will be positively ecstatic. It's just one hell of a gamble...

**Update: I added a table thanks to a suggestion by TheJay and took players similar in size/position to Jennings from this 2000+ shot list (first 3 yrs), and then compared to the total career percentages. Here's what I found.

Loose comparison of first three years (2000+ shot attempts) vs. career shooting percentages
Player Initial Attempts FG% Career Attempts FG% Change
Jason Williams "White Chocolate" 2280 38.40% 7699 39.80% 1.40%
Jason Kidd 2679 38.70% 14941 40.10% 1.40%
Raymond Felton 2904 39.60% 6012 41.10% 0.50%
Larry Hughes 2366 39.60% 9189 40.60% 1.00%
Nick Van Exel 3106 41.00% 11458 39.50% -0.50%
Mookie Blaylock 3106 41.20% 11499 40.90% -0.30%
Damon Stoudamire 3627 41.20% 10662 40.60% -0.60%
Stephon Marbury 2991 41.70% 13324 43.30% 1.60%
T.J. Ford 2089 41.80% 4215 43.30% 1.50%
Gilbert Arenas 2478 42.00% 8786 42.10% 0.10%
Randy Foye 2134 42.00% 3381 41.20% -0.80%
Baron Davis 2731 42.10% 11767 41.00% -1.10%
Vernon Maxwell 2701 42.10% 9878 39.80% -2.30%
Jason Terry 3186 43.00% 12947 44.80% 1.80%
Allen Iverson 3967 43.10% 19906 42.50% -0.6
Ben Gordon 3546 43.10% 7579 43.50% 0.40%