Posted by Raphy on October 6, 2011
The postseason is off to an exciting start with 3 series reaching a dramatic sudden death game five. Here are the years with the most sudden death League Division Series games:
This entry was posted on Thursday, October 6th, 2011 at 5:23 am and is filed under Game Finders, Postseason.
You can follow any responses to this entry through the RSS 2.0 feed.
Both comments and pings are currently closed.
October 6th, 2011 at 6:08 am
Wow... and there were none before 1981. I think the LDS has seen their numbers increase due to "Performance enhancers."
and by that I mean guys like Griffey Jr, Grace, Jeter and Schilling.
October 6th, 2011 at 6:27 am
@1
The only year before 1995 which had an LDS was 1981.
October 6th, 2011 at 7:45 am
2001,2002,2003 were the only 3 years out of 16 with more than one sudden death after the strike.
October 6th, 2011 at 7:46 am
@2 and @1 - For example in 1984 there was no LDS, the Cubs played the Padres in the NLCS and the winner went to the World Series. 1981 was a strike year so there was a playoff.
October 6th, 2011 at 8:00 am
Heh... '81, '01, & '11 ...of course, '91 didn't have an LDS, but it's interesting all the sudden death LDS's are in the xxx1 part of the decades so far. I guess that means we'll have 3 exciting LDS's again in 2021? haha
October 6th, 2011 at 9:04 am
1981 was 1st half winner against 2nd half winner. Current format is in it's 17th year.
October 6th, 2011 at 9:08 am
Considering how the regular season ended, I guess 2011 will be remembered as the season of the wires. "Coming down to the wire" described two Wild-Card runs, and now two LDS's.
Makes me wonder what the league serieses {seriesi??} and World Series will bring.
October 6th, 2011 at 9:11 am
I have to say I was not a fan of the idea of the LDS, but I was sold the minute I saw Griffey slide home and have that huge grin on his face at the bottom of the pile.
October 6th, 2011 at 9:26 am
Seems like LDS should last the full five games a lot more often than they have historically. From 1995-2010 you only had 14 LDS out of a possible 64 that went the distance (22%.)
If all postseason wins and losses were random coin flips, then about 37% of all best-of-five series should last five games.
If you try to compensate for home-field advantage (the home team in games 1-2 is a combined 292-225, .565 in postseason history, and the home team in games 3-4 is a combined 251-220, .533) then you still end up with about a 38% chance of the series going five games.
Probability that you'd get 14 or fewer heads out of 64 flips of a coin that has a 38% chance of landing heads = 0.45% (or you could say the probability that a .380 hitter gets 14 or fewer hits in a stretch of 64 at bats = 0.45%.)
What am I missing here?
October 6th, 2011 at 9:28 am
Amendment to #7;
I meant 3 LDS's {or, is that LDSii?}
Sorry 'bout that...my granddaughter is working on her freshman grammar {and driving Grandma and I nutzoid in the process}.
October 6th, 2011 at 9:36 am
@6
Current format is in its 18th year. This is the 17th in which games were played using it.
October 6th, 2011 at 10:12 am
@11: 18th year but no wild card in 1981
October 6th, 2011 at 10:25 am
@9, Cloycebox -- I haven't looked at the numbers, but it could be partly due to:
(a) Unevenly matched teams.
(b) The top 2 teams in each league are more likely to have clinched in time to set up their playoff rotation. The wild card often has to go all-out right down to the final few games.
October 6th, 2011 at 11:03 am
@9: The players only get 60% of the gate on first three games of the LDS series.
Nothing for a game 4 or 5. Don't think they get a penny of TV revenues. TV revenues all go to regular season contracts. That could go along way in explaining the bending of the probabilities. 22% seems awful low.
October 6th, 2011 at 11:22 am
"Current format" meaning 3 divisions, 3 champs + wild card. The format officially began in 1994, but the strike wiped out not only the last third of the season but the entire postseason.
1981 was the abortion of the after-the-fact split season, with the first round of the improvised two-round playoffs pitting the winners of the first half and the second half in each division (along with the further abortion that if a team won both halves, it would NOT get a bye, but would play the second-place team with, I guess, the best record, I forget). Let's hear a big cheer for Bowie Kuhn. Anyway, '81 was NOT in the current format, even though it did feature an additional round of playoffs.
October 6th, 2011 at 12:12 pm
The first 3 years of the LDS (1995-1997), it was a 2-3 format with the series starting at the home of the lower seed. In that period, 2 of 12 series went the distance, about 17%.
Since then, it has been the current 2-2-1 format, with the series starting and ending at the home of the higher seed. In this format, it's 18 of 56 series going the distance, about 32%, or almost twice as frequently.
My guess is the current format is more likely to promote a limit series since the lower seed is perhaps more likely to be trailing after the first 2or 3 games (this year, lower seed trailed 3 of the 4 series after 3 games), but gets to try to make a comeback at home. Which is kind of ironic, as the shift to the current format was intended to help the higher seed by getting to start the series at home.
October 6th, 2011 at 12:26 pm
@15.
Further '81 anomaly. The Cardinals had the best record in the NL over the entire '81 season, but missed the post-season as they finished 2nd in both halves of the season.
October 6th, 2011 at 12:32 pm
@17.
I should have said the Cardinals had the best record in the NL East over the entire '81 season.
Another anomaly - teams ended up playing quite different numbers of games. In the NL the range was 102 games for the Cards and Pirates, up to 111 games for the Giants. In the AL, it was 103 for the Indians and Royals, up to 110 games for the Angels.
October 6th, 2011 at 1:11 pm
@9
If you want to get really technical, instead of using a coin flip, maybe use the teams' in-season records. So if a team with a .600 winning percentage plays a team with a .550 winning percentage, the chance of the .600 team winning is about 55.1%.
Or better yet, maybe use the seasonal runs scored vs. runs allowed for each team. Then you calculate for each series the likelihood of it going 5 games, and see whether this matches reality.
If not for my insufferable laziness, I'd run the numbers.
October 6th, 2011 at 1:14 pm
@18
The Reds had the best record in all of baseball in 1981, but didn't make the playoffs either. It should have been the Cards vs. the Reds. After that season the Reds traded George Foster and Ken Griffey, Sr., and the dismantling of the Big Red Machine was basically complete. Bench and Concepción were still there, but no one was around them.
October 6th, 2011 at 1:15 pm
Doug (16) makes a good point -- I assumed that every playoff series in MLB history had games 1&2 in the ballpark of the team with home field advantage, and games 3&4 in the other team's ballpark. But there are a few historical exceptions to that rule that I could have carved out. Probably wouldn't have changed much.
Let's say you create a series that looks really lopsided though -- suppose the team with home-field advantage has a 70% chance of winning a game in their own park and a 60% chance of winning on the road. There should STILL be about a 31% chance of a series like that lasting five games.
October 6th, 2011 at 1:37 pm
@16: 15 of 56 about 27% since 2-2-1 format.
12 of 52 before this year 23%.
October 6th, 2011 at 4:28 pm
Some more numbers to consider:
-- 41 of 106 World Series (38.5%) were tied 2-2 at some point.
-- from 1969-84, the LCS were best of five, and 10 of 32 (31%) needed a game 5.
-- from 1985-2010, the LCS were best of seven, and 13 out of 50 (26%) were tied 2-2 at some point.
-- 17 of 68 LDS (25%) have gone to a game 5.
-- all best of five series in MLB history (excl 1981): 27 out of 100 had a game 5.
If you flip four coins, the probability of getting exactly two heads is 37.5%. That's very close to the World Series number. As you go back to the earlier rounds, you're probably more likely to get mismatches than in the World Series. So all of that makes sense to me.
But I still think it's very surprising how rare it is to have a game 5 out of 5. If you get a LDS between two teams, one of which has a 70% chance of winning each game, you still have a 26% chance of the series going five games. You wouldn't expect any playoff team to win 70% vs any other playoff team -- 70% is Phillies vs Astros, not Phillies vs Cardinals.
October 6th, 2011 at 7:30 pm
@23:Like your original assessment at 38%. Mismatches don't always come in to the winner's circle. The 1954 Indians & the1990 Oakland A's both got swept 4-0 in the World Series.
October 6th, 2011 at 9:08 pm
#20.
Don't forget about Bench beginning to really slow down by that time as well, and playing more third base {which I have to admit, he did pretty well} than behind the plate. Add to that the fact that Cincinnati is a comparatively small-market town to have a team in the Free-Agent Era, and you can see why the Reds have been a generally disappointing franchaise since '80.
October 7th, 2011 at 4:13 am
@21.
Re: Games 1 and 2 on home field of lower seed.
This didn't just happen the first 3 years of the LDS. The first 16 years of the LCS (1969 to 1984) had the 2-3 format until the LCS was extended to a best of 7 series in 1985.
October 7th, 2011 at 6:16 am
@9: Stan Musial's 1948 .376 batting title is almost exactly on your
coin flip analogy. (6-16)
His average dropped from .411 on May 20th to .367 on September 8th. He may of had a 12 for 52 stretch during that period. (.230)
October 7th, 2011 at 6:26 am
Coin flips:1 of 16 =4 heads.
1of 16 =4 tails.
3 of 16 = 3 heads.
3 of 16 = 3 tails.
4 of 16 = 4 heads.
4 of 16 = 4 tails.
6 of 16 = 3 heads or 3 tails. (.375)
October 7th, 2011 at 6:38 am
@28: Screwed that coin flips up pretty good.
How does that go again? It's been a while since I played a 4 team parlay.
October 7th, 2011 at 7:30 am
4 team parlay 16 combinations.
4 winners =1
4 losers =1
3 winners =4
3 losers =4
2 winners =6