clock menu more-arrow no yes

Filed under:

Warriors coasting so hard in the regular season, it forced 538 to revise their model

New, comments

“The Warriors would lose games that ultimately didn’t matter,“ say the predictive modelling experts at FiveThirtyEight.

Photo by Ezra Shaw/Getty Images

As the Golden State Warriors enter the halfway mark of the season and find themselves not holding on to the top seed in the West, some fans may be getting nervous. After all, the 16th-rated defense is not going to cut it...right?

After winning a record 73 games three seasons ago, but ultimately losing in the NBA Finals, the Warriors have slowly backed off their win total, with a firmer eye on being ready for the Playoffs. Last season, the Warriors finished without holding the top seed for the first time since this run began - but they “flipped the switch” in the playoffs and played the best defense in the league, in spite of going up against the NBA’s top offenses.

That “switch” makes everything harder to predict

This has been a problem for a while for me personally, trying to talk about LeBron James’ defense. You see, in the regular season, his defense is legitimately abysmal - and yet when he turns it on in the playoffs he can be one of the toughest in the league. So when someone asks “is LeBron a good defender” it really throws me for a loop because it’s almost like two separate evaluations. I end up hemming and hawing and calling him bad in the regular season, but good in the playoffs - but I still don’t really have a good answer.

FiveThirtyEight (which I’ll just call by their common nomenclature “538”) is known for their accurate political polls, but they’ve also dipped their toes deep into other realms of statistical analysis. Their predictive basketball models may not be 100% accurate, but it’s something that they are continually working on an improving.

They just announced a major new update to how their models will be presented, and they used our Warriors as their example to explain why.

An evolving model - with some serious problems

So, here’s the issue the modelling experts ran into - and it is also at the heart of why many of the GSOM writers have been discounting concerns about the Warriors poor record so far this season:

During the regular season, the Golden State Warriors would either hit dry spells while key players were injured, or coach Steve Kerr, guarding against injuries, would frequently sit key players. Because of that, the Warriors would lose games that ultimately didn’t matter that much, and our Elo model — designed to track the ebbs and flows of the season’s wins and losses — would punish Golden State each time. But like clockwork, the Warriors would emerge in the spring as strong as ever — despite their blemished Elo rating.

Wait. So you’re telling me that a meaningless loss to say, the Los Angeles Lakers on Christmas day doesn’t mean that the Warriors aren’t going to make the playoffs?

Using the old model, 538 says that they’ve got Golden State with a “head-scratching” 27 percent probability of reaching the NBA Finals and a 14 percent chance of winning their third consecutive title.

The new model? Much more in line with reality:

AH, that looks a little more accurate

So there you have it, advanced stats say the Warriors are extremely good, and we shouldn’t be overly concerned with a subpar regular season. Be sure to go over to 538 and check out their full post, as it gives a lot more details, and you’ll want to bookmark their new pages if that’s your thing.

For now though, it sure is nice having some statistical certainty that further reinforces my belief that the Warriors are too good for us to read too much into some of these regular season struggles.