The NBA playoffs start Saturday, April 15th, and the prohibitive favorites to win the title are the Golden State Warriors at an average price just north of -200. South Point has them at -220, Westgate is at -200, MGM has -200, and William Hill has them at a hefty -250. These numbers would suggest that the Warriors have about a 65% chance to win it all this year. That leaves 35% worth of probability for the rest of the league (15 remaining playoff teams) to chop up. The ESPN BPI (Basketball Power Index) which is their oft-cited analytics formula, has the Warriors at 71% to win the title while at 538.com they have them at 59%. These numbers suggest that these prices of -200 and higher are warranted.
The point of this discussion is to try to illustrate that when one team is so dominant in the betting market it’s almost impossible to make the rest of the field add up to construct a proper futures market using traditional hold percentages. There’s just not much left to do aside from making everyone else an average of 40-1.
Let’s extend the premise of the books’ numbers roughly aligning with the analytic projections. You’d have the Spurs as second choice. Analytics say +750, while the books have differing prices on the Spurs (+450 MGM, Westgate +800, and offshores at +1000). The average is in the +750 range, so they are in the same ballpark there. After San Antonio the odds are long on the analytics side, with the Raptors the third choice at a 4.2% win probability, which equates with a line in the 20-1 range. Actual lines are more like 40-1. Analytics place another half dozen teams in the 2%-3% range, including the Rockets, Clippers, Celtics, Wizards, and Jazz. The sixth team in this group is the Cavaliers, who had a rough week heading into the playoffs. They lost a home game to a Hawks team that was resting four of its five starters, then a few days later blew a 26-point fourth-quarter lead in Atlanta. The Cavs rested players as the season wound down and even when they didn’t, it looked as if the starters were just going through the motions. As a result, their stock dropped considerably according to the stat guys and what once was a team given a 20%+ chance to win the title dropped to around 3% on the spreadsheets.
Should the Cavaliers really be priced at 30-1 like the non-gambling media projections suggest? ESPNs BPI says they have a 3.3% probability of winning the title, while 538 gives them just a 2% shot. Those percentages certainly suggest that 30-1 would be a reasonable price. A melding of the prices from those two analytic sources would come up with a line of about 36-1 on the Cavs. Discount the price for the house hold and 30-1 is about what it would be if the books were roughly using the probabilities being dished by the mainstream analytics. But they’re not.
The Cavs can be had in the 3-1 to 5-1 price range in the books, so the Cavs line is where the analytics and the books part ways. Analytics are not supposed to be affected by public perception and are meant to present a probability based on set performance factors and projected success rates vs. upcoming opposition. Can they factor in whether or not a team was playing up to its capabilities or take into account that the personnel used in the most recent outings were replacement players? Weighting these non-statistical elements is not where a raw number crunching approach shines. Adjustments based on more than numbers have to be made.
The Cavs’ drop-off in projected win probability was due to the team’s performance in the last month of the season. The current state of the NBA is one where resting players, tanking, and jockeying for playoff-seeding position in the final week of the season have all become commonplace. Can these analytical projections be trusted, or is it the books that offer a more realistic assessment of a teams chance of success. ESPN and 538 don’t book bets and that’s a good thing for them. But if they wanted to MAKE a bet on what their numbers suggest is an overlay, then 5 Dimes and a few other books are offering the NO on the Cavs winning the title at -480.