Did the win last night doom the Celtics' lottery odds? Well, not really.
First of all, if you are tempted not to believe me, read this e-mail I sent on March 8th:
|Yeah, some redaction is in order.|
I said the above after a surprising win against the Nets, and when the Celtics beat the Pistons the night after, I surely sounded like a silly optimistic fan wearing tank-colored glasses. Yet as things stand now, the Jazz have won the same number of games as the Celtics have, and I'm one loss away from being more clairvoyant than Bill Simmons is.
So, why did I say that with a seeming level of confidence? Well, if you read my posts on this very blog, you know that I play around with numbers as much as I can, and even though just looking at the numbers cannot paint you a very good picture, it gives you a very good idea about how things can turn out. To quote a famous statistician: "All models are wrong, but some are useful." If you come up with a model that can approximate reality, then the odds of your predictions coming true will be considerably high.
Here's a basic assumption that holds more or less: For a team's every surprising win, there's a surprising loss proportional to its winning percentage. When the Celtics won against the Heat, we shouldn't have had any more wins for a while against good teams, and a surprising loss would have been in order: Thus we lost to the 76ers. Again, there are cases when teams get hot for a while and that changes everything, especially in close situations like this, but it is more likely that such a thing shouldn't happen twice in the same season. The Celtics, the Jazz, the 76ers, the Lakers etc. have all had their runs for 2013-2014.
Let's cut to the chase: Where do the Celtics (24-55) stand as of today? Well, if we miraculously win all of our remaining games, we will have the same number of ping-pong balls as the Kings (27-53). That is really unlikely, but if it happens, the Celtics' probability of having a top-3 pick will be 0.18, and we'll most likely pick at #7. Not pleasant, I know.
Here's the critical question: What if we win two of our last three games? Are we doomed then? Well, not exactly. See, the Lakers (25-54) are going to play the Jazz (24-55) this week, and the last time I checked, someone has to win that game. Unlike the Celtics' hypothetical wins against the Cavaliers and the 76ers, that win is pretty real. So, if the Lakers win, that will put them at 26 wins, and we will have the same numbers of wins. That'll mean a 0.25 chance at having a top-3 pick. If the Jazz win and lose the rest of their games, that will put us at #6 with odds at 0.21. What is funny is that that's not a drastic difference compared to our worst case scenario, given the depth of this draft.
What is the likeliest scenario? Given the current injuries to Sullinger (ankle), Bayless (knee), Humphries (knee), Rondo (shin) (remember what I said about "phantom injuries" in that e-mail? Don't be surprised when at least half of our injured list is given a chance to rest for any given game) the Celtics will be able to pull at most one more win. This is where things get -seemingly- complicated, depending on whether the Lakers or the Jazz win that crucial matchup, yet in any case, the Celtics will either rank #5 or share odds with the aforementioned teams, which will more or less mean the same thing. Our odds of having a top 3 pick will then be ~0.29.
Oh, and of course there's the dream scenario that the Celtics lose all of their remaining games, and they will either have the 4th best odds or share the odds with the Jazz. The expected value of a top-3 pick in that scenario is ~0.35.
|Dreaaaaam ooooon, dreeaaaaaam oooooon...|
But again, all models are wrong. Don't blame me if the Celtics do something really stupid in the last three games.