Hollinger's Playoff odds

by Replacement Level Poster on Dec 23, 2007 1:58 AM PST Comment 7 comments

http://sports.espn.go.com/nba/hollinger/playoffodds

Above is think to the Hollinger playoff odds, and for the first time in just about forever, we are actually projected to make the playoffs. Granted our odds are 50.2% so it is basically a coin flip at this point.

Still, I bought season tickets with a co-worker when we got the number one pick. Then Oden got hurt, and we thought "Oh well, its fun to go to games, this is our investment in next year". I didn't think we'd be at this point yet, and maybe in two weeks we won't be at this point, right now though it sure is fun!

0 recs | Comment 7 comments

Story-email Email Printer Print

Comments

I think you got your money's Worth

and more than expectedd!

Hold onto your seats. I am waiting to see what happens with the next game. Christmas Day Red VS Green. I'm expecting a WIN. Bring 'em on.

One step at a time. Closer & closer we get to the predictions!

"Lenny Suckerpunch NEVER bet on me!" - Elizabeth "The Lizzard" Lowblow

by Lizzy Lowblow on Dec 23, 2007 3:29 AM PST reply actions 0 recs

I'll take that as SOME validation...

...of my current inclination to think it more reasonable than not to believe they'll make it this year (though, as you sort of note, being so close to 50%, the technically most epistemically justified (there's a bit of a philosophical term for any philosophers out there!) or rational position to take on the basis of that evidence would be suspension of judgment, i.e., "I have no clue either way at the moment".

I just wrote some extended thoughts on why I'm at where I'm at in a reply to Dave and tssbro in the diary "So what would it take, Dave?" (for anyone that has the graciousness and actually gives a rip to care about my thoughts!), so I won't elaborate on that here again. But to whatever extent Hollinger's system/process for determining likely playoff teams has a rhyme or reason to it, I'll just reaffirm that I'm seeing what it is "seeing"; there's something going on here with this team that's making it not unreasonable to say, "Huh. Maybe it's not so crazy to think..."

But in the end, I'm the type that thinks it most important for people to believe what appears most reasonable to them to believe, not what some random dude like me thinks, and to enjoy the present ride along the way for everything it's worth. And since I don't want to be responsible for excess disappointment on the part of others, maybe I should shut up on all of this. (Dealing with my own disappointment - which would be minimal regardless, frankly - I can handle.) But if the Blazers come out of their next extended road trip winning the majority of the games on it, God help me.

When feeling discouraged about the Blazers' learning process, just remember that at least the Jail Blazer era is over.

by kdfkwak on Dec 23, 2007 4:18 AM PST reply actions 0 recs

Does not compute

In the NBA, either a team makes the playoffs, or it is in the lottery. Either/or. There is no middle ground. Therefore the sum of the percentage chance that Portland will be in the playoff plus the chance that it will be in the lottery must equal 100.0.

Hollinger estimates that the chance that Portland will make the playoffs is 50.2 percent. Necessarily, therefore, the chance that Portland would be in the lottery must be 49.8 percent. But Hollinger states the odds of Portland being in the lottery as 0.1 percent. Impossible.

When a clock strikes 13 times, you do not say, "Oh, it's thirteen o'clock!" you say, "The clock is broken." Hollinger's odds do not compute.

by monkeysuncle on Dec 23, 2007 5:52 AM PST reply actions 0 recs

my guess is

those are projected odds of winning the lottery...you're absolutely right about what the odds of being in the lottery are, and thus i doubt he felt the need to include them...otherwise it wouldnt make sense

since he has the knicks at a 0 percent chance to make the playoffs i'd guess they have more than a 19 percent chance to get a lottery pick

by DominicanAvenger on Dec 23, 2007 6:49 AM PST up reply actions 0 recs

ahh yes, from the bottom of the page

"Lottery = Likelihood of winning the lottery"

by DominicanAvenger on Dec 23, 2007 6:51 AM PST up reply actions 0 recs

Holllinger's Odds, Implications for the Blazers

A few comments on Hollinger's Methodology:

The way that Hollinger calculates his "playoff odds" (really playoff probabilities) is kinda cool, but certainly has some limitations. As he says on the page, he calculates the odds by taking the current record and then using his power ratings to simulate the rest of the season 5,000 times. Because the higher rated team sometimes loses, each time the computer simulates the season, it will generate different records each time. Hollinger's play-off odds are just the proportion of times each team made the playoffs in the 5,000 simulated seasons.

The cool thing about this method is that it takes into account who each team has beat thus far and by what margin as well as its schedule in the rest of the season. So, if a team has a good record simply because it has had an easy schedule, then it will be unlikely to win at it's current rate.

The bad thing about this methodology is that it almost certainly over-estimates the probability that the good teams make the playoffs and that the bad teams miss the playoffs. Why? Because it assumes that the current power ratings are the true power ratings (that is has been estimated exactly) and that the current power ratings will hold throughout the rest of the season. Both assumptions are problematic and as a result there are too many teams in his ranking that have either a 100 percent chance of making the playoffs or a 0 percent chance of making the playoffs.

Most obviously, future power ratings can change because of injury. If the Pistons lost a couple of starters to injury, there is a good chance that they wouldn't make the playoffs, but Hollinger gives them a 100 percent chance of doing so. Similarly, one of the reasons for the Heat's poor record is Wade's injury. I think they have a better than 6 or 7 percent chance of making the playoffs, considering the 8th place team in the East has a losing record.

How could his system be improved? One strategy would be to use past seasons to estimate the probability that a team with a given record makes the playoffs. Unfortunately, there aren't a lot of seasons with the current number of teams and play-off spots. On the other hand, even 10 seasons of data would be enough to show that Hollinger's estimates are not "spread out" enough.

What does all of that mean for the Blazers? Well, that depends on several things. Most importantly it dependents on whether or not you think that the teams' current win streak means that they are better than their overall record. If so, then the chance that the Blazers make the playoffs is certainly higher than 50%. By contrast, if you think the Blazers are injury prone and inconsistent (due to their youth), then you might make an argument that their chance of making the playoffs is less than 50%.

In my view, the fact that the Blazers have played so much better since James Jones joined the active roster suggests that the Blazers really are better than their overall record. He's not the Blazers best player, but he's added enough to make the Blazers better and more consistent. I also think that Outlaw is more than a player on a hot streak. The coaching staff has figure out how to use him and he is logging more minutes than he was to start the season. He is simply a different player than he was when the season started. Those two factors together are enough to make me think that the Blazers have a good shot at that 8th spot. If I were betting money, I'd take odds/payout that gave the Blazers a 65% chance of making the playoffs.

I give the Blazers a 35% chance of missing the playoffs because a) an injury to Roy could happen and, b) Houston, Golden State, and Utah can all play much better.

by PoliSam on Dec 23, 2007 1:00 PM PST reply actions 0 recs