Better thinking through technology

The role of attractors in basketball betting

Every sporting event has two strong attractors that pull it in opposite directions. Unsurprisingly, these two attractors represent how much each team is expected to try to pull the game toward their side of the spread. Basketball, however, has the unusual distinction of being a game where the score is constantly being refreshed. That is to say that the score in basketball is a constant reminder of how well the two teams are playing. As evidenced by the number of games where two teams have close rebounding and TO figures AND the score is close, we can reasonably assume that the score reflects a fairly honest telling of how teams are playing in terms of the traditional Four Factors model.

Because basketball games go back and forth with possessions that come in quick succession, basketball game spreads oscillate more obviously and clearly than spreads do in other sports. The back and forth of a game is made readily apparent simply by looking at the score. Likewise, we could easily check the stats box for rebounds, TOs and shooting figures and quickly guess the score even if we didn't know it.

The challenge in trying to configure a live betting strategy for any basketball game is determining just when the game should be reasonably expected to swing. This is why it is important to determine what the attractors are for each game.

Let's take a look at a pre-game set of predictions made for Game One of the LAC v HOU series from the 2015 playoffs.

The prediction was that that this game could be reasonably expected to wobble between a HOU -14.5 lead and a LAC -13.5 lead.

Three things need to be kept in mind here.

First: Some games get caught in a loop on one side of the graph. The CHI@CLE Game One did this, with CLE never realistically able to assume a lead.

Second: A game can slingshot past the attractor points pretty far. For example, the CHI largest lead went well past the projected CHI -5.5 lead. In fact, the biggest Bulls lead of that game ended up being an early 16-point lead, and the Cavs never led at all.

What we can say for reasonably sure is that, if both teams are expected to continue to fight to take the lead, that 16-point Cavs lead should swing back toward the -5.5 attractor and try to push back toward the projected middle (for that game it was a -4.5, which was in fact the prevailing pre-match spread at most sports books following the announcement that Kevin Love wouldn't play).

Third: The critical juncture is as a game whips from one side toward the center.

As we saw in the CLEvCHI game, just because a team fights its way back across the opponent's attractor and toward their own doesn't mean that they'll magically get to their side of the graph. The Bulls would be expected to fight back and try to whip the game back toward their own attractor.

In this regard, a bad NBA game can end up being similar to getting caught in a circulating current near the shoreline. You may fight to get back to the surface, but if you fail, the current will drag you back toward (and potentially past) a predictable location in space.

As for the LAC@HOU game, here's the map of how the game circulated toward both attractor points . . .

If you actually check the play-by-play, you'll see that the Rockets did try, unsuccessfully, to pull the middle game back toward their attractor during the 3rd quarter. The Clippers, however, were able to breakhtrough and dragged the game past their own attractor, with a largest lead of 18 points and a final spread (and win) of 16 points.

It's especially important in live betting to avoid getting caught on the wrong side of the current of a game. This goes double if a game is approaching its end.

For example, the CLE game went well beyond the CHI -5.5 attractor, but that lead for the Bulls hit early enough that bettors could reliably purchase a higher spread bet in live betting and expect the game to contract toward the middle attractor (the CLE -4.5, in that particular case).

Likewise, the LAC +13 for the game would have been good value in 2Q because we could reasonably expect the Clippers to drag that game back toward their side of the bet.

Where things get tricky is when teams lose the will to keep fighting. The Bucks' elimination game against the Bulls is a good example. The attractor point for that game was in the mid-teens, but the Bucks simply stopped trying to pull the game their way, and the Bulls just kept building and building a lead until the very end.

It's important to pay attention to a game as it approaches the attractors and check for evidence that a team is fighting to stay in the game. Offensive rebounds matter a lot. Also, a team should be avoiding dumb turnovers (stepped out of bounds, for example). Once more, this goes double in elimination games, and just double the fear factor again when you're looking at weaker teams like the Bucks and the Nets, who may just decide to end their seasons an hour or two early if the better team takes a solid lead against them.