Does it Matter Who You Play?

Posted on April 16, 2008 by


Editors Note: Today’s guest column is authored by Jason Eshleman.  For regular readers of the WoW Journal, this is the “Jason” who offers so many wonderful comments and who authored a guest column last December.  In addition to posting frequently at the WoW Journal, Jason is a research associate in Anthropology at UC Davis.  Jason co-founded Trace Genetics, which is described on their website as “specialists in genetic identity DNA analysis with expertise in DNA ancestry, forensics, ancient DNA analysis, molecular diagnostics, and population genetics.” When Jason is not doing stuff that is far more complicated than discussing the intricacies of the NBA, he’s living life as a long suffering fan of the Golden State Warriors.

A season comes to an end

A 122 to 116 loss sealed the deal.  Despite 48 wins (with a chance for 49 remaining) the Warriors would miss the playoffs.  Warriors out, Denver in.  Though it was still possible that Golden State could finish the season with a record identical to the Nuggets, Denver — by virtue of established tiebreakers –had sealed the 8th spot.  The key tiebreaker: Denver has won 3 of 4 matches between the two clubs. 

This fact suggests the obvious:  Denver held their advantage in the standings by virtue of their head to head record.  The corollary? Against the rest of the league, the Warriors had the superior winning percentage. 

To read Warrior message boards and blogs, there is presently no bigger tragedy in the world than Cleveland, Washington, Toronto, Philadelphia and Atlanta all seeing their seasons continue while all finishing with records inferior to the Oakland team’s winning percentage.  The Hawks (and possibly the 76ers) will even finish with a losing record.  Rising unemployment, home foreclosures, a faltering economy and $4/gallon gas have nothing on this injustice.  “The playoff format must be revised! It should be the best 16 teams, no matter what.  This just isn’t fair!” 

No, it isn’t fair.  Life, as many have noted for many a year, is not fair.  (Foul poles are fair, though baseball is a peculiar sport.)  Inferior teams will have an opportunity to play basketball while the Warriors go on an early vacation.  It is all an accident of having to play against better teams simply because of geography.

Not.  Fair.

East vs. West

While many have regarded the Eastern Conference of the NBA to be the weaker conference for several years now, this year the relative strength of the two conferences has seemed especially disparate.  Teams in the east have been less successful (at least as assessed by wins and losses) than teams in the west.  All totaled, western conference teams have a winning record while the eastern conference as a whole has a losing record.  Not even the strength of the Celtics and Pistons, who will finish with the league’s best records, changes this.

This fact suggests the obvious: since intra-conference games produce as many wins as losses for that conference, the Western conference teams must have this winning record by virtue of victories in head-to-head match-ups with Eastern conference teams.

The disparity in conference strength is not new, nor has the league ignored it in formulating various rules for post-season play.  To guard against teams beating up on weaker non-conference rivals, following head-to-head record, intra-conference records break ties when teams finish with identical records.  Going into their head-to-head contest in Oakland on the 10th of April, Denver had already secured any possible tiebreakers over the Warriors. In the event of a tie, they would hold the key “better conference record.”  Against teams in the West, Denver has been the more successful ball club. 

The corollary: At the risk of sounding like the bitter bloggers-and allow me this late disclaimer: I am a Warriors fan and have been for many, many bitter years- this of course means that against the Eastern Conference, against supposedly weaker opponents, Denver could not match the Warriors’ record!  Prior to their head to head match up, David Stern was prepared to crown Denver as playoff worthy even though they hadn’t been able to beat those weaker foes in the East as often as the Warriors had!  How could he possibly favor one team over another when that team could not keep pace with their rival against the lesser clubs? We have got to end this! It is not fair that teams should be rewarded simply for beating up on inferior opponents!

This is, of course, absurd. These methods, placing rigid guidelines that are inevitably rather arbitrary in their logic, are, in fact, absurd. So is concluding incontrovertible superiority on the basis of a one-game advantage, the difference in inter-and intra conference records between the Nuggets and Warriors. 

But absurd lines of reasoning (and disregard for statistical reality) have not prevented some from suggesting that Lebron James is not worthy of the MVP award merely because of the record of his opponents. Despite statistical performances that suggest he should at the least be a leading candidate, some  have suggested that his stats are at least in part a result of playing an unbalanced schedule against weaker foes (and of course Kobe Bryant suggests himself that he would look better if he played 52 games against eastern conference squads though actual results do not reflect this bias);

But is this real?  Do players in the East have inflated stats simply by being the best of a bad lot?  And do teams in the West sport the records they do because they’ve faced stronger conference foes? Has this imbalance determined who was in and who was out of the playoffs?  Are some teams better at picking on the weak?  Here, numbers do provide an answer.

Winning teams win, losing teams lose

Over an 82 game season, teams play 52 contests against the 14 other teams in their conference and another 30 games against teams from the other conference.  It’s hypothetically possible that a team could finish with a superior record by merely holding their own against their own conference and piling up non-conference victories.  Split in conference but pound the lesser squads and you can finish with a winning record.  It could happen. 

But it didn’t.  It’s a strategy employed with successful results by exactly zero teams in the NBA. When we look at the records of all 30 clubs – noted in Table One — and compare their overall winning percentages to their conference winning percentage, there was not a single club that had an overall winning record that did not also have a winning record in their own conference. Losing clubs almost invariably (Sacramento being the lone exception) have to date had losing records out of conference as well. 

Table One: NBA Records with One Day Left

Not a single team had a statistically significant difference in their in-conference and overall record.

It’s important to note that whenever we look at numbers in such a manner, some variation is to be expected simply due to the sample.  52 games is not necessarily enough to conclude that a 0.600 winning percentage really means a club was better than a 0.650 winning percentage over 32 games. But we can control for this by estimating the probability that both represent subsets of the same sample to see if there is any significance to the difference. And there wasn’t. Differences between overall record and conference records were but a matter of the noise associated with sample size, and in these cases, the differences were quite small indeed.

To summarize: Winning teams are winning teams.  Losing teams are losing teams, and the conference schedules seem to provide little influence on who was good and who was bad.  In general, Boston was good regardless of geography and Seattle stunk no matter the opponent.  A team’s record is a good, perhaps not perfect, but still very good gauge of the team’s quality.

All of this of course suggests that it is indeed unfair that the Warriors will once again watch the playoffs on TV. 

– Jason Eshleman

Posted in: Uncategorized