More Sweeps = More Quitters?

May 3, 2009 2:47 PM

Are there more sweeps and/or fewer comebacks in pro sports playoff series these days, or is it just my imagination?

I've been following pro sports for more than 20 years, and, while there have always been 4-game sweeps in the playoffs (the NHL featured four straight in the finals in the 90s, though at least two were outright mismatches), it seems as if there are more 4 and 5-game series these days than there used to be. Or, at the very least, it seems that teams that fall behind in their series rarely ever come back and win. I'm talking especially about the past 3-4 years here (and I don't follow the NBA much, so my discussion will be limited to the NHL and MLB).

A perfect illustration of what I'm talking about is the Colorado Rockies' 2007 playoff run. The club went on a tear late in the season to make it into the playoffs, and thereby developed this weird "mystique" that allowed them to sweep their way into the World Series. Clubs that should have at least had a chance against them (particularly the Diamondbacks) couldn't even win a game. Yet as soon as they got into the Series and got blown out in Game 1 against the Red Sox, they themselves folded the tent and got swept out.

Really, I can't think of a really compelling MLB playoff series since the 2004 ALCS and NLCS. The 2008 Series was awful, in 2006 the Tigers cruised into the Series and then suddenly forgot how to pitch against a rather unimpressive Cardinals club. Series that should be competitve simply weren't.

As for the NHL, the playoffs seem to have lost the magic they once had. Even though the 90s were plagued by Finals sweeps, the 80s and 90s overall featured compelling rivalries, ranging from Edmonton-Calgary to Detroit-Colorado. The first round this year included three 4-game sweeps, and only one compelling series (Capitals-Rangers). Again, it seems as if teams, when faced with adversity, simply give up.

I'm not sure if there are numbers to back me up, but I guess the two stats that I would be interested in knowing about are: 1) Are there more 4 and 5-game series these days than there were, say, 15-20 years ago?, and 2) Do more clubs that go up 1-0 in playoff series go on to win than they used to 15-20 years ago?

Moreover, if anyone agrees with me, do they dare tie this into sociological trends concerning Gen-Yers' inability to stick with things, tough out tough circumstances, etc.?

posted by hiteleven to Sports, Hobbies, & Recreation (12 answers total)

No. You are relying on anecdotal evidence, which is never the answer. You are mixing in arbitrary delineations (why point out only Finals sweeps, or include nonsense categories like "compelling series?"). Then, assuming all of your random observations are meaningful, you're jumping to #138 in the list of likely reasons for these occurrences. If you actually do take the time to crunch the numbers, which is easy in this internet age, you will find little change once you reach levels of statistical significance. Hopefully, that will lead you to read fewer silly Gen-Y stereotypes (really?) into the state of sports.
posted by aswego at 3:27 PM on May 3, 2009 [2 favorites]

I know you said you don't follow the NBA, but most commentators are claiming the Celtics-Bulls series that just ended which went seven games was one of the best ever. And in the first round of the NHL playoffs, both the Devils-Canes and the Rangers-Caps series went seven. And even in the 06 MLB playoffs, the Mets-Cardinals NLCS went seven.
posted by hobgadling at 3:31 PM on May 3, 2009

In the NBA Atlanta and Miami just finished a 7 game series. The closest game was won by 10 points. Are you saying that was a close series because it went 7 games?

I did go back and check how each NBA series would have ended under the 5 game series length (which was extended to 7 games a few years ago). All had the winner of the 7 game series also winning 3 games first.

As for the Rockies example that you cite, I would bet that it has something to do TV scheduling. The Rockies ended their Divisional Series game on October 6 and started the NLCS 5 days later. Going 5 days without playing can be a big deal when that doesn't happen during the season, something that was somewhat masked by Arizona being in the same situation date wise. The Rockies then won the NLCS on October 15 and didn't play again until October 24. Meanwhile, Boston ended their ALCS on October 21. 9 days off versus 3 can be a big deal especially in a game like baseball where having a rhythm is so important.

I'm not saying that these scheduled off days have everything to do with it. But these kind of delays are a modern phenomenon.

Also, a significant portion of current professional athletes are not part of Generation Y as defined by Wikipedia (born 1981-2001).
posted by theichibun at 3:32 PM on May 3, 2009

I looked at the Stanley Cup Playoffs from last year (2008) and compared it to twenty years ago (1988) with respect to your two specific questions. Last year had exactly one fewer series where the team that lost the first game went on to win the series (4-3). On the other hand, last year had two fewer series that lasted only four or five games (8-6). I don't cite this because it "proves" your theory wrong. It really doesn't, as those are arbitrary choices of seasons and aren't large sample sizes. Nevertheless, as data goes, they're a hundred times more valuable than the random memories that happened to stick with you that led you to develop your theory. Those are extremely valuable for reminiscing about exciting moments and drawing enjoyment from the game, but they are shit when it comes to actual analysis.
posted by aswego at 3:42 PM on May 3, 2009

The first round of the NBA playoffs used to be 5 games, now it's 7. So naturally, longer series these days.
posted by drjimmy11 at 3:56 PM on May 3, 2009

Response by poster: I know you said you don't follow the NBA, but most commentators are claiming the Celtics-Bulls series that just ended which went seven games was one of the best ever.

If I got a nickel every time some hyperbolic sportscaster said that about this or that sporting event these days, well, you know the rest...


Also, a significant portion of current professional athletes are not part of Generation Y as defined by Wikipedia (born 1981-2001).

Most pro athletes hit peak performance in their mid to late-20s, fitting easily within your age range (which is just one of many aribtrary choices, by the way...Wikipedia is not gospel).

Also, I'd prefer more answers that actually cited numbers (whether proving me right or wrong), as opposed to those from slighted Gen Yers (I threw that bit in as a bit of a joke, anyway). I realize I'm citing anecdotal evidence, but simply denying my claim because of that is not a convincing counter-argument, either.
posted by hiteleven at 4:20 PM on May 3, 2009

Best answer: For baseball the answer is yes

I took the data from this site, divided it by decades (OK, that is arbitrary) and came up with games per 7 game series. Now if you really care, we are skipping the 5 game series (I believe 1970 era playoffs), and the 9 game series (like 1919)

The data shows, to me, a peak from 1940-1980

http://www.whowins.com/tables/all.html

1900 5.4
1910 5.4
1920 5.5
1930 5.4
1940 6
1950 6.1
1960 6
1970 6.1
1980 5.9
1990 5.7
2000 5.6

I will have hockey and basketball momentarily
posted by rakish_yet_centered at 4:56 PM on May 3, 2009

Best answer: For hockey, It is valid to say that series have become longer

1939-49 5.45
1950 5.47
1960 5.55
1970 5.31
1980 5.60
1990 5.70
2000 5.72

As for basketball, 7 game series have become shorter

1947-1959 6.00
1960 5.97
1970 5.87
1980 5.50
1990 5.66
2000 5.64
posted by rakish_yet_centered at 5:18 PM on May 3, 2009 [1 favorite]

You're correct. It's a combination 1) less "gamers" (gritty players who will do anything to win) in U.S. pro sports and 2) greed.

These (North) American leagues keep adding more playoff games and rounds because it means more revenue. And they space the games so far apart now that by the time the finals roll around, most of the participants are too worn out (or disinterested) to play at full tilt and they get sloppy (especially in the NBA).

Once one of the teams gets an edge, the other team is more apt to throw in the towel (especially in the NBA).
posted by Zambrano at 6:06 PM on May 3, 2009

Response by poster: Thanks for the link and for the number-crunching. Some interesting results, particularly for baseball...the dropoff from the 60s-80s to today is significant enough, I'd say.

To look at some more granular data, here are the number of sweeps and 7-game series for each playoff year in the NHL since 1990 (the same number of playoff rounds have been played every year, for the non-fans).

1990 4-game series: 2 7-game series: 4
1991 4-game series: 0 7-game series: 4
1992 4-game series: 5 7-game series: 6
1993 4-game series: 2 7-game series: 4
1994 4-game series: 2 7-game series: 7
1995 4-game series: 4 7-game series: 4
1996 4-game series: 2 7-game series: 2
1997 4-game series: 2 7-game series: 3
1998 4-game series: 3 7-game series: 1
1999 4-game series: 3 7-game series: 3
2000 4-game series: 2 7-game series: 3
2001 4-game series: 3 7-game series: 4
2002 4-game series: 0 7-game series: 5
2003 4-game series: 2 7-game series: 6
2004 4-game series: 1 7-game series: 5
2006 4-game series: 2 7-game series: 3
2007 4-game series: 1 7-game series: 1
2008 4-game series: 2 7-game series: 3

All told it looks like the best runs were in the early 90s and 2000s. Anyway, interesting numbers to look at.
posted by hiteleven at 6:13 PM on May 3, 2009

Let's objectify this a little bit. Instead of studying for finals, I used Excel to run a few regression analyses on the data at the above link (http://www.whowins.com/tables/all.html). Of course, I am not a pro statistician by any means; I am just ridiculously familiar with the process of running this analysis from the research project (conservation of an endangered Florida bird) that I've been working on for the past year and a half. Statistics experts might disagree with the use of this test.

It's important to note that this analysis has to do with trends in number of games. This has nothing to do with number of sweeps or number of 7-game series specifically. It answers the question, "Are series being decided quicker now than they used to be?" This is not the same as the questions posed by hiteleven, which are "Are there more sweeps than there used to be?" and "Are there fewer 7-game series than there used to be?" (I decided against answering those directly because it's a lot more work to answer those, and I really do need to study at some point.) This analysis quantifies and analyzes the statistical significance of the general decadal trends noted by rakish_yet_centered.

MLB: It appears that there is no long-term trend in number of games played in the MLB (p = 0.63), although the 1960s peak pointed out by hiteleven is apparent from a 2nd-degree polynomial trendline plotted on a scatter plot graph of (year of series) vs. (number of games in series), where each year can have more than one series. (Note that I have no idea how to test for a "peak.").

NBA: Using p = 0.05 as the level for significance, there is a negative trend in the NBA (p = 0.017), apparently confirming hiteleven's hypothesis.

NHL: The trend is actually positive (p = 0.044); that is, the trend is towards more games in each series (which would imply fewer sweeps and more 7-game series).

There is, of course, the question of whether the difference observed is practically significant. Is a difference of an average of 0.36 games per series (as seen in basketball between 1947-1959 and the 2000s so far) really significant to us as sports fans? That's a subjective question.
posted by dondiego87 at 9:49 PM on May 3, 2009

d87, as you might have guessed, I am not a mathematician, I decided that the number of sweeps wasn't important, because the number of series per year has increased, especially in hockey and basketball, I guess I could have done sweep percentage, but it exaggerates the arbitrary decade division. Let me show you a chart...Matt, where is the img tag?...oh, nevermind

I think all 3 sports have changed enough that to come to a conclusion based upon this data is ridiculous. Take baseball, one thing this is measuring is the strength of the strongest team to the playoff team that they play. Well, if there is only 1 series per year, you presumably get the two strongest teams, better chance of a long series. When you add playoff teams, now #1 might be playing #8, greater chance of a sweep. In hockey and basketball, since they went to 16 team playoffs, even more so.

I think the data shows that it's a complex analysis since the 3 sets of data have gone in different directions.

As far as measuring character of the modern athlete.... please.
posted by rakish_yet_centered at 5:21 PM on May 4, 2009