So . . . the NBA season opened a couple of days ago and we're pretty much ready to crown the champions. After a 1-1 record, losing to heavyweight Boston (literally, now that they have Shaq and Jermaine O'Neal) and beating lightweight Philadelphia (whose new coach is already missing games and blaming it on a concussion from months ago), the oddsmakers, pundits, and so-called experts claim that the Miami superfriends are ready to be champions. Based on odds I spent all of about 4 minutes researching, they have a 35-40% chance to win a title. More on the problems with these odds after the jump(-ball).
Let's start with the odds; a team deemed to have a 35% chance of winning the title has about a ten times greater chance than is their rightful due (30 teams in the league = 3.3% chance for a league with perfect parity). And Miami is loaded with talent, at least in 3 of their 15 roster spots. The defending-champion Lakers are given the second-highest odds of winning the finals, with about a 25% chance. This means that if you were to only bet on these two teams, you have about two in three chance of having bet on a winner. Put another way, Miami and LA have sewn up two thirds of the odds of winning, and the other 28 teams share the last third.
(Extended aside: both of the sports betting sites from which I drew the odds have superadditivity problems, which means that if you add up all of the odds for all 30 teams they total greater than 1. Given that, to my knowledge, There Can Be Only One champion, there's a mistake somewhere. So, if you have $3000 lying around, take it to the bookies and bet $100 on each team. On one of the sites I visited, this would result in a guaranteed profit of $1200! Unless there is something I don't know about sports betting, which is possible. I do know about odds, however, and odds should add up to one.)
If this were April and we were down to 16 teams, I might buy that two teams dominate the odds in this way; given that it is October, and the season has just started, I ain't buying it. If I had money to bet, I would bet on the field. I may lose, but I firmly believe that there is just too much variability to claim that one team is such an overwhelming favorite. If the NBA was played as a computer simulation (isn't that what they do with the BCS?) and injuries, suspensions, trades, and other relatively low-probability events were removed, I'd be right on the Miami/LA bandwagon. But that's not how the real world works.
As Nassim Nicholas Taleb talks about in the Black Swan, those events that have the lowest probability (and therefore are often ignored) tend to have the greatest impact (see: housing crisis/crash of 2009). It is highly unlikely that Dwyane Wade will be arrested and incarcerated (maybe for consistently misspelling his name), but if he were it would have a severe impact on his team (as for Kobe being arrested, well . . . let's not go down that path).
It is somewhat likely, however, that one of the Miami superfriends (LeBron, Wade, Bosh) gets injured, and that would have an extremely detrimental impact on the team (same with Kobe in LA). I would argue that Miami (and to a lesser extent, the Lakers) have a greater risk of negative outcomes due to injuries than do other teams, because there is a greater disparity between the star players and the rest of the team. If one of the Raptors gets injured, it's not going to have that bad an impact (when you're only going to win 15 out of 82 games, is there still room for a negative impact). Even a team that is considered somewhat good, like Atlanta or Boston, is not going to see their fortunes fall as far due to injury as a team that relies on one or two stars. And because Miami has three essential players (and assuming injuries are random), they have a 20% chance of a key player being injured (3 of 15 players on the roster), whereas a team with one star player has a 6.7% chance of a bad injury befalling a critical piece of the roster.
Should be an interesting ride. And Miami and LA certainly each have a good chance of being champions, better than any one other team. But I'm not sure the chances is greater than all the other teams.
I agree with your general idea, but I believe your logic is flawed in your "likelihood of a key player being injured" near the end...
ReplyDeleteThe percentages you are giving are the chances that, if a player is injured, that player is one of the key players, not actually the chances that a key player gets injured.
In fact, I suspect that, in general, the chances of a key player being injured are actually a bit higher than that, because a) most teams have more than one player injured in any given season and b) key players play more minutes and are therefore, presumably, more likely than average to be injured.
On the other hand, in order to have a substantial impact on their likelihood of winning a title, an injury to one of the Heat's key players would have to be severe enough or occur late enough to cause him to miss games in the playoffs. (I am assuming they would make the playoffs even without one of the key guys, especially in the East.)