Fantasy Blogs  |  The Court Jester  |  

If you’re as obsessed with fantasy hoops as I am, you may not have noticed that something odd’s going on. But I’ve been discovering some weird coincidences. Certain of my players always seem to play on the same night. When guys go down injured there’s a very limited subset of players whose minutes increase. And the nightly box scores are grouped in such a way that certain players’ names always appear in close proximity. I’ve recently figured out that this is because there is another kind of basketball competition, besides fantasy, going on. It’s called the “real NBA season.”

The NBA is apparently a keeper league that combines features of the auction and the snake drafts. As in fantasy hoops, team owners attempt to assemble the most talent, but the prize pool is considerably higher than typical leagues, with billions of dollars changing hands instead of a couple of hundred and maybe a pizza thrown in for draft day lunch.

Among the other oddities are that the players on a given squad all reside in the same city and wear cute little matching suits. Evidence of how simplistic and unsatisfying this competition is is that the only category that counts is points.

Though I’m not really sure what the allure is to a true fantasy junky, apparently there are people out there who do care, because thousands of them pack arenas to watch despite fact that the stadium box score screens don’t typically list steals or blocks.

Apparently there is an “NBA playoffs” which comes after the fantasy playoffs. I’ve never watched, but it is probably interesting as a pre-preseason league where you can scout talent for the following fall’s fantasy draft.  

But still, despite their irrelevance, it might be worth looking at NBA team dynamics to see if they yield any clues that might be useful in fantasy.

First off, here’s an answer to an interesting question. Using the fantasy metrics, what real team generates the best stats? If you were drafting blind, which team would you want a guy from? Let’s say each team’s numbers were boiled down into a single player, who plays about 36 minutes. What would his stats be?

Across teams, the “typical” 36 minute fantasy player has pretty similar numbers in most categories – about 15 points, 6.2 boards, 3.3 assists, 1.2 steals, 0.8 blocks, and 2.1 turnovers. The major variation comes in three point shooting (which ranges from the Magic’s 1.5 to several teams’ 0.6), field goal (a spread from Phoenix’s .489 to the Nets’ .425), and free throw (Dallas’ .819 to Detroit’s pathetic .715).

It’s a lot easier to make sense of these numbers by looking at standardized scores, which judge the teams relative to each other based on standard deviations above or below the mean. Note that a high score in turnovers is bad. The stats are:

As you can see, Dallas is the top team in the league at producing fantasy numbers, thanks to great free throw shooting, strong assists, very low turnovers, good blocks and basic competence at the other categories. Also way out ahead of the pack is Phoenix, with stellar field goal, points and threes, plus good assists. Perhaps it’s no coincidence that Nash and Kidd, crafty old floor generals, quarterback these stat-stuffing squads.

However, Phoenix falls down badly in steals. That’s not a problem for Denver, which also has strong scoring. The Lakers (excellent rebounding) and Hawks (stunningly low turnovers) round out the top five.

Which teams are fantasy graveyards? The Wizards (terrible assists and steals), the Bobcats (all-round mediocrity), the T-wolves (zero blocks, sky-high turnovers), and way in the rear the Nets (gaping holes at field goal, threes, points, rebounds, and assists) and the Pistons (well below average in every category except turnovers).

It’s certainly surprising to see there’s a team that’s worse than the Nets, but unsurprisingly, on the whole winning teams put up winning fantasy numbers. This is borne out by the correlation coefficient, a number that measures the strength of the relationship between two statistics. A correlation score of 1.0 means that when one number is high the other one is always high too. A 0.0 means the two stats have no relation, and a negative number meaning they tend to move in opposite directions. The correlation between a team’s winning percentage and the fantasy numbers it generates is 0.759, a very strong relationship. In real life good fantasy numbers generally translate into lots of wins.

This is pretty good evidence that the basic formula on which we fantasy players depend is pretty sound. We hope that our nine cat system is not just an arbitrary statistical construct but is truly a good way to measure players’ ability to win basketball games. From these numbers it appears it is.

But is this true for all of the categories? Which individual categories do the best job of predicting whether real life teams win basketball games? Here are the correlations:

Field goal is the statistic most strongly associated with winning, even more than points. Assists are very important, and threes, blocks, and rebounds matter too, though less so. You’ll note that turnovers are a pretty important category for winning, so those of you who for some bizarre reason play in 8 cat leagues should wake up and add this important and relevant stat.

Perhaps most surprising is that based on this year’s numbers steals and free throw shooting have practically zero relationship with winning real basketball games. I’m not recommending we change our game, but it might be worth doing a larger study to see if these categories really contribute to wins. If not, theoretically they should be dropped from fantasy consideration.

Though interesting, these numbers have limited relevance as far as being a guide to winning at fantasy. The direction of causation certainly mostly goes from strong stats to winning and not vice versa, so seeing one of your guys traded from a bad team to a good one will, all else equal, probably do little for his numbers and may hurt them due to increased competition for minutes. (Though I am starting to wonder if Steve Nash might really help the offensive numbers of his teammates.)

But in doing this analysis, I found something which definitely piqued my interest as far as fantasy strategy, overturning one of my most zealously held drafting precepts. More on this next time.