Biggest NBA Finals Collapses, 1992-2010
Posted by Neil Paine on May 31, 2011
With Dallas-Miami Part II tipping off tonight (not that it's really a rematch), I wanted to see whether the Mavs' loss in 2006 was the worst Finals collapse of the BBR era. We have linescores for every playoff game since 1992, which means I can calculate the home team's probability of winning at various checkpoints within a game:
| Stage | p(Home W) |
|---|---|
| Pregame | 60.4% |
| After 1st Qtr | =1/(1+EXP(-0.3599755-0.1122741*Home Margin)) |
| After 2nd Qtr | =1/(1+EXP(-0.2895922-0.1429087*Home Margin)) |
| After 3rd Qtr | =1/(1+EXP(-0.2041572-0.2117494*Home Margin)) |
| Before any OT | 52.4% |
Combining those probabilities with the series win probabilities I found here, one can determine each team's probability of winning the series at a given checkpoint. This allows us to rank Finals collapses, pinpointing the moments within games where the eventual loser's series win probability was the highest:
1. Dallas loses to Miami, 2006
Dallas' highest series win probability: 90%, after 3rd Quarter of Game 3 (Mavs led series 2-0, game by 9 pts)
Miami Series Win Probability Graph:
2. Utah loses to Chicago, 1998
Utah's highest series win probability: 69%, after 3rd Quarter of Game 2 (Jazz led series 1-0, game by 3 pts)
Chicago Series Win Probability Graph:
3. Philadelphia loses to Los Angeles, 2001
Philly's highest series win probability: 66%, after OT of Game 1 (76ers led series 1-0)
Lakers Series Win Probability Graph:
4. Boston loses to Los Angeles, 2010
Boston's highest series win probability: 66%, after 1st Quarter of Game 7 (series tied 3-3, Celtics led game by 9 pts)
Lakers Series Win Probability Graph:
5. New York loses to Houston, 1994
New York's highest series win probability: 65%, after 1st Quarter of Game 6 (Knicks led series 3-2, game tied)
Houston Series Win Probability Graph:
Worst of the Rest:
6. Orlando loses to Houston, 1995
Orlando's highest series win probability: 62%, after 2nd Quarter of Game 1 (Magic led game by 11 pts)
7. New Jersey loses to San Antonio, 2003
New Jersey's highest series win probability: 59%, after 1st Quarter of Game 3 (series tied 1-1, Nets led game by 6 pts)
8. Utah loses to Chicago, 1997
Utah's highest series win probability: 57%, after 1st Quarter of Game 5 (series tied 2-2, Jazz led game by 13 pts)
9. Portland loses to Chicago, 1992
Portland's highest series win probability: 54%, after OT of Game 2 (series tied 1-1)
10. Los Angeles loses to Detroit, 2004
Los Angeles' highest series win probability: 53%, after 2nd Quarter of Game 1 (Lakers led game by 1 pt)
(Click here for raw quarter-by-quarter series win probability data for every Finals since 1992.)
May 31st, 2011 at 9:39 am
This list should have been count down from 5 to 1, but then again as soon as I read the oost title I knew #1 was 2006.
Fan up!
May 31st, 2011 at 11:39 am
I'm a little surprised that the 10th biggest upset of the past 28 years involved overcoming "just" a 53% deficit. Given that that was achieved via a 1 point lead in the 2nd quarter of game 1, it seems that many series were won by teams that took early leads and held them. Am I missing something?
May 31st, 2011 at 2:13 pm
Hey, what about counting the series-long probability with the area under the line?
May 31st, 2011 at 2:39 pm
"Am I missing something?"
There wasn't simply that many series to choose from?
May 31st, 2011 at 3:19 pm
BSK, come to think of it, there really just haven't been that many great finals. (I've been watching them for 20 years.) Even #3 on the list really overstates things since it was clearly a mismatch - I wonder what the Vegas series line was after Philly stole game one.
May 31st, 2011 at 3:34 pm
#2 - Like WiLQ said, there are only 19 Finals to choose from here to begin with (we're only going back to 1992). Also, the team with home-court won 14 of them. If you have HCA, you start out with a built-in 53% shot before the ball is tipped, and it takes a lot to move the needle very far early in the series.
Then, in the middle of the series, the 2-3-2 takes over, as the win probability formula knows that the HCA team has 2 back-to-back home games in their back pocket. You would think that '97 Utah, up 13 in Game 5 of a 2-2 series, would have a higher series win expectancy than '03 New Jersey, up 6 in Game 3 of a 1-1 series, especially since neither had HCA. But the chance of the home team in Game 3 (tied 1-1) winning a 2-3-2 series is 54%; the same probability for the home team in Game 5 of a 2-2 series is 45%. The final 2 games loom larger in Game 5 than in Game 3.
That's why if the Mavs win, it will likely be in 5 games -- if it gets to a Game 6 or 7, Miami's chances go up considerably because of the HCA.
May 31st, 2011 at 9:28 pm
i think the dallas collapse in 2006 was the biggest i have seen. mavs up 2-0 and double digit lead in 4th quarter of game 3 before falling in 4 straight. of course the refs had a lot to do with it too - 46 FTs for wade in games 5 & 6, danny crawford anyone - but the mavs let it slip away. too bad nash didnt stay with dallas or they would have won 1 or 2 by now. probably should have won it in 2003 when they started out 15-0, but injuries to finley and dirk grounded them in the WCF.
boston collapsed badly in the 2nd half of game 7 last year too. you could see that one coming in LA though. celtics started settling for long 3's and got tired with a short bench (perkins injury) and the intensity of the game. wallace started out well but ran out of gas. LA limped to the title last year with bryant shooting poorly in game 7 and playing on a bad knee. it isnt surprising they came up way short this year. they were fortunate to win it all in 2010.
the 2009 orlando finals was much closer than 4-1 as well. magic could easily have been up 3-1 with just a few plays being different. they collapsed pretty badly in game 4 down the stretch when it should have been tied up 2-2 instead of going down 3-1. howard missed key FTs and they let fisher shoot an open 3 when orlando was up 3 in the final seconds. dumb, blew a five or seven point lead in the final minute-plus.
utah collapsed badly in game 6 of the 98 finals too. orlando as well in game 1 of the 1995 finals. who can forget nick anderson missing those 4 FTs late as orlando blew a big lead to houston, lost in OT and then got swept.
boston had a big collapse in game 4 of the 1987 finals, but bad refereeing and a lack of using the bench had a lot do with that one.
if dallas wins one of the first two in the 2011 Finals, i like their chances. if you get a split early having the next 3 at home usually equals at least a 3-2 lead. last year it didnt mean a title for boston, but with a healthy perkins i think they win game 7.
May 31st, 2011 at 9:31 pm
Yea, I realized that we had a limited data set. It just seems weird to consider a series an "upset" because the winning team trailed by 1 point halfway through game 1. I guess that was solely the function of a road team winning.
June 1st, 2011 at 3:27 am
ben
out of the NBA finals i have seen myself, going back to before age 10, these are the best in my opinion:
15) 2000 LA 4, Indiana 2
14) 1992 Chi 4, Ptld 2
13) 2005 SA 4, Det 3
12) 1991 Chi 4, LA 1
11) 2010 LA 4, Bos 3
10) 1993 Chi 4, Phx 2
9) 1988 LA 4, Det 3
8) 1998 Chi 4, Utah 2
7) 1980 LA 4, Phi 2
6) 1987 LA 4, Bos 2
5) 1977 Portland 4, Philly 2
4) 1976 Bos 4, Phx 2
3) 1997 Chi 4, Utah 2
2) 1974 Bos 4, Milw 3
1) 1984 Bos 4, LA 3
There really havent been many good Finals this century. 2010 was close and hard-fought but the quality of play was uneven. 3 of the 4 Spur titles were one-sided against inferior teams. Only the 2005 series vs. Detroit was good and close.
Worst Finals
1) 2007 SA 4, Cle 0
2) 1999 SA 4, NY 1
3) 2002 LA 4, NJ 0
4) 2001 LA 4, Philly 1
5) 2003 SA 4, NJ 2
6) 1979 Sea 4, Wash 1
Biggest Upsets
1) 1975 GSt 4, Wash 0
2) 2004 Det 4, LA 1
3) 1977 Ptld 4, Philly 2
The best playoff series I saw was the 1981 East Finals between Boston and Philly.
Best 1st rounder, especially since they went to best 4 of 7, was the 2009 Chicago/Boston 7-game classic with 4 OT games. In best 3 of 5, I thought the 1999 1st round series where Utah edged the Kings 3-2 was a great one. The 1984 1st rounder between Detroit and the Knicks also was a classic, as was the 1991 Indiana/Boston series.
A really good 2nd round series was the 1980 WCSF where defending champ Seattle edged Milwaukee in 7 games.
June 3rd, 2011 at 11:08 am
Ah what a surprise, Celtic Finals are your favorite.
Not really making you look objective and I don't think the quality of league in the 70's was comparable either. ABA sucking up some of our superstars, whereas now every bonafide star plays in the NBA.
Europe has some role players but not on the level of a prime Dr. J, Rick Barry, Gilmore.
June 3rd, 2011 at 11:35 am
Watching the Bulls is far more entertaining than the other 4-2 series you have on your list. Especially against a mediocre defensive team like the Suns in 1993.
If you want another close series you can move up the 98 series against Utah.
June 4th, 2011 at 4:53 pm
I don't know the metric used here, but it sure like there was only one collapse that ever happened. In both #2 and #3, the team that won was the team everyone expected to win, and all they did was lose the first game.
Even after Philly won game 1 in 2001, it was hard to imagine them beating the Lakers. That was a David and Goliath final if I ever saw one.
June 13th, 2011 at 8:20 am
So it looks like this year's "collapse" would rank 2nd on the list (Miami was almost at 80% after the 3rd quarter of game 4), how fitting.