October 1, 2006
Locking it Up
Does Clinching Early Help?
Drive a nail home and clinch it so faithfully that you can wake up in the night and think of your work with satisfaction, a work at which you would not be ashamed to invoke the Muse.
New York is abuzz with the prospect of another Subway Series, and who can blame them? Though both New York teams took very different routes--the Mets dominated the ignoble National League all season, while the Yankees needed a late-season surge and a five-game sweep of the rival Red Sox to take charge of the AL East--they both were the first in their respective leagues to clinch a playoff spot.
On September 18, the Mets shut out the Marlins, 4-0, and had a 91-58 record, a 14 1/2-game lead, and 13 games left in the season, thereby clinching their first NL East title in 18 seasons.
One day later, the Yankees clinched the AL East title for the ninth straight season with a 3-2 win over the Blue Jays. It has become so commonplace that all the media could do is goggle at the Yankees actually welcoming supposed pariah Alex Rodriguez to their celebration. The Yanks had a 92-59 record, an 11-game lead, and 11 games left in their season.
The Mets became just the 39th team in baseball history to clinch a division or league title with at least 13 games remaining in the season and the first since the 105-win 2004 Cardinals. Keep in mind that the Mets had to keep their champagne on ice during a three-game sweep at the hands of the lowly Pirates. The Mets could have been just the 17th team in baseball history to clinch a title with 16 or more games remaining (the last being the 2002 Braves).
Meanwhile the Yankees became the 75th team ever to clinch a title with 11 or more games left to play. That's the third most games that they have had in hand while guzzling the champagne for their current title run. They had 12 games remaining when they clinched in 2001 and 19 in their epic 1998 season. In all three cases, the Red Sox were the last team eliminated. (By the way, the earliest the Yankee franchise has ever clinched was in 1941: with 20 games remaining, they eliminated the White Sox.)
There are four other races, however, that are or were at the other end of the spectrum. The NL wild card and NL West were nominally decided yesterday, as the Padres and Dodgers locked up postseason spots. Which team wins the division and which is the wild card will be determined today.
The other two races have been anything but expected. On August 7, the Tigers owned the best record in baseball (76-36, .679) and a 10-game lead in the AL Central race. Today, the last day of the season, they are tied with the Twins. Detroit has not spent a day out of first since before May 16. Minnesota had never been in first until Thursday night. The race is somewhat blunted because both have clinched a playoff spot, but both teams would prefer to play the first two games at home than in the Bronx in the first round. The Twins, nabbing at least a share of first place for the first time this late in the season, become the second latest in "modern" baseball history to do so, and missed the record by one day.
The current modern "record" is held by the Cardinals who tied the Reds for control of the National League pennant on September 29, 1964 en route to a National League pennant and World Series crown. The Reds took the league lead just two days earlier from the infamous 1964 Phils, who were up by 6 1/2 games on September 17 but lost 12 of their last 15 games. (The all-time "record" is held by the 1873 Boston Red Stockings, who took over first for the first time on October 1, though the season lasted another month.)
Which brings me to the NL Central. This season the Cardinals are the ones with the big lead and long losing streak (which is gratifying for a Phillies fan like me), eight of their last 10 games. St. Louis was seven games ahead of Cincinnati and 8 1/2 games ahead of Houston on September 17. Since then they are 3-9, and their lead is down to a game-and-a-half over the Astros. It does not hurt that the Houston, who was last in first place on April 27, won nine straight before arriving in Atlanta Friday.
All three of these races went down to the last weekend of the season. Given that the two New York clubs clinched very early and the final three spots will be clinched very late in the season (while the A's clinched the AL West title with five games to go), does that give those two New York teams an almost insurmountable edge heading into the playoffs?
Looking at all first place teams throughout baseball history, we can divide them into three almost equal groups of early clinchers, later clinchers, and average clinchers. Early clinchers eliminated all their opponents with at least eight games to play. Late clinchers had no more than three games remaining. That leaves average clinchers with between four and seven games remaining.
The Yankees and Mets clearly fall in the early clinching group, the A's fall into the average group, and whoever wins the final three divisions will likely fall into the late-clinching group. (We will leave the wild cards for later.) By our theory above, the early clinchers (Yankees and Mets) will have a distinct advantage over the teams that will clinch later. Let's look at our clinching groups in baseball's past to determine if that is true.
First, let's examine the basic statistics for each group:
Clinch Type Avg PCT Avg W Avg L PCT Avg GA Games Remaining Early .647 98.23 54.11 .645 12.18 11.91 Avg .611 89.60 58.48 .605 5.81 5.43 Late .606 90.86 60.31 .601 2.82 1.48 Overall .622 93.05 57.66 .617 6.88 6.18
Obviously the early group has by far the best overall record (.647 winning percentage). However, the early and the late group are separated by just five percentage points. Again, it appears that the early clinchers are head and shoulders above the other two groups.
However, before we proceed I want to go back and include that pimple on the face of the playoff system, wild card teams, in the analysis.
Determining when wild cards clinch is somewhat problematic. In some cases they are still in contention for their division title after clinching a playoff spot. Actually, the wild card also throws a monkey wrench in analyzing certain pennant races, those in which both teams have already clinched a playoff spot and in which becoming the division winner is almost academic. In my analysis, I have decoupled the wild card races and division races, as if the wild card were a separate division altogether.
The statistics above for wild card teams are as follows:
Clinch Type Avg PCT Avg W Avg L PCT Avg GA Games Remaining Early .611 99.00 63.00 .611 14.50 10.00 Avg .583 94.50 67.50 .583 6.50 4.25 Late .568 90.81 69.00 .568 2.28 1.13 Overall .575 92.23 68.18 .575 4.16 2.50
They are considerably worse teams than the division winners above. However, the early group is clearly superior to the other two groups.
Now here are the stats for both groups combined (first-place teams and wild cards):
Clinch Type Avg PCT Avg W Avg L PCT Avg GA Games Remaining Early .646 98.25 54.25 .644 12.22 11.88 Avg .610 89.79 58.84 .604 5.84 5.39 Late .602 90.86 61.26 .597 2.76 1.44 Overall .613 93.00 58.29 .615 6.72 5.96
How have those teams fared in the postseason? Here are the results:
Clinch Type #Tms Won Div? Won Lg? Won WS? % Early 122 53 87 40 33% Avg 101 47 72 32 32% Late 147 62 91 33 22% Overall 370 162 250 105 28%
The late clinching group does win a much smaller percentage than the other two as expected. However, the early clinches fare about as well as the average group even though they have a much better winning percentage.
If we look at the winning percentage in playoff games, the outlook gets even bleaker for early clinchers:
Clinch Type PostW PostL PCT Early 400 370 .519 Avg 355 310 .534 Late 411 481 .461 Overall 1166 1161 .501
(Note: The five games missing in the loss total belong to the 1892 Cleveland Spiders, who were a second-place team that got swept in a "World's Championship Series" against the regular season champs, the Boston Beaneaters.)
The early group may have played more games, but their winning percentage is worse than the average group.
Could clinching early and playing a larger number of games that "don't matter" cause a team to get rusty and have a sub-par performance in the postseason? Perhaps early clinchers go into a spiral in the inconsequential games after eliminating their foes. Below are the records of each group after they had clinched a title:
Clinch Type PCT W After L After PCT Diff Early .646 779 565 .580 -.066 Avg .610 245 235 .510 -.100 Late .602 79 88 .473 -.129 Overall .613 1103 888 .554 -.059
The early clinchers not only have the best winning percentage, they perform much closer to their overall performance than the other two groups. This is all the more remarkable given that they play many, many more of these pointless games. The average group is barely above .500 and the late group has a losing record.
This does not appear to support the "rust" theory. Then again, we are not looking at postseason performance here. Could it be that this "rustiness" sets in only when the postseason commences? Can we see this rust settling on the early clinchers in some sort of statistical means?
What does having more time to rest players mean to a team? Well, resting starting pitchers means more time to set up the most advantageous pitching rotation. One would expect that perhaps over-tinkered pitching staffs would show some signs of sluggishness in the postseason, at least more so than the offenses would. Let us test that theory by comparing the overall earned run averages during the regular season and the postseason per clinching type:
Clinch Type Regular Season ERA Postseason ERA % Change Early 3.29 3.31 0.34% Avg 3.49 3.48 -0.32% Late 3.54 3.63 2.48% Overall 3.44 3.48 0.98%
Yes, there is a very slight ERA increase for the early group's pitchers, and a very slight improvement in the average group's, but the early clinchers still have a better overall ERA. There is very little indication of pitching staffs being overly rusty.
Actually, the one thing this comparison bares out is that the late clinchers, the teams with little or no time to set their rotations and rest their key pitchers, are the only teams that are negatively impacted, and it sure isn't rust that causes their woes.
If rust has little effect on pitching, perhaps the position players are the ones that are afflicted. Let's compare the offensive ratios per group for the regular season and postseason:
Clinch Type RegOBP RegSLG RegOPS PostOBP PostSLG PostOPS OPS % Diff Early .341 .410 751 .311 .366 676 -9.89% Avg .336 .405 741 .315 .372 687 -7.30% Late .336 .407 743 .311 .370 681 -8.34% Overall .338 .407 745 .312 .369 681 -8.56%
It appears that pitching does, as the aphorism goes, dominate the postseason. All of the groups saw an appreciable decrease in their offensive numbers. However, the early clinchers get hurt the most while the average group sees the least impact.
Apparently, the rust theory has some merit, but unexpectedly, it's the well-rested position players that show the ill affects.
But why does it afflict the early clinchers but leave the average group largely unaffected? Could it be that having a handful of meaningless games at the end of the season provides the average group with some much-needed rest while the early clinchers, with almost an extra week to rest, begin to calcify with the aches and pains of the offseason? A little rest is a help, but a lot of rest is a hindrance? It appears to be the case. Perhaps that's why it affects the pitchers much less given that an extra week off would mean perhaps one lost start rather than an extended idle period.
So what does this all mean? A New York Subway Series is definitely a possibility, but it is by no means the lock that it would appear to be. The Mets swoon since clinching (3-7) would seem to be foreshadowing this.
Or are they? Is there a relationship between a team's record after clinching and how well they perform in the postseason? I ran the numbers and there is very little correlation between a team's winning percentage after clinching and their winning percentage in the postseason (correlation coefficient of 0.112).
However, maybe winning percentage does not capture overall performance especially when late clinchers play perhaps just a game or two. I approached it differently: I categorized teams based on winning (.500 or better) or losing record (sub-.500) or no games remaining in the season. I ran the postseason numbers for all three groups:
After Cat Post W Post L PCT Avg PCT # Tms WS Won % Losing 395 398 .498 .439 104 32 31% Winning 650 603 .519 .476 179 62 35% No Games 121 160 .431 .360 46 11 24% Overall 1166 1161 .501 .448 329 105 32%
It is preferable to continue to win after clinching, but it clearly is more important to have at least a handful of games to prepare for the postseason.
So although it would be better for the Mets to win once in a while even if the games no longer really matter, but maybe they should worry more about who is replacing Pedro Martinez.
Finally, for the karmically inclined fans of "My Name is Earl," let's take a look at whether what a team does on the day they clinch has any bearing on their postseason fate. Does "backing in" to the playoffs (i.e., losing a game and still clinching) hurt a team's chances in the postseason?
Well, I looked it every which way that I could and I saw very little indication that winning on the day that you clinch matters once the postseason rolls around. One complication is that some teams played a doubleheader on the day they clinched, some played one game, and some did not play at all.
There was almost no correlation a) between the team's winning percentage on their clinching day and whether they will win the World Series (0.012 coefficient), b) between the number of wins on a team's clinching day and whether they will win the World Series (also 0.012), c) between postseason, winning percentage and winning percentage on the day a team clinches (0.031), and d) between the number of wins on a team's clinching day and postseason winning percentage (0.022).
Evidently, what a team does in order to clinch a spot-win, lose, draw, or rest-does not foreshadow one whit what they will do in the postseason.
For the record, here are the teams that clinched first place with the most games remaining.
Yr Tm Lg Div ClinchDate LastTmElim Games Remaining W L PCT 1902 PIT NL 19020903 BSN 25 103 36 .741 1999 CLE AL C 19990908 CHA 23 97 65 .599 1995 CLE AL C 19950908 KCA 22 100 44 .694 2002 ATL NL E 20020909 PHI 21 101 59 .631 1941 NYA AL 19410904 CHA 20 101 53 .656 1975 CIN NL W 19750907 LAN 20 108 54 .667 1943 SLN NL 19430918 CIN 19 105 49 .682 1998 NYA AL E 19980909 BOS 19 114 48 .704 1887 SL4 AA 18870915 LS2 18 95 40 .704 1904 NY1 NL 19040921 CHN 17 106 47 .693 1936 NYA AL 19360908 DET 17 102 51 .667 1986 NYN NL E 19860917 PHI 17 108 54 .667
Here are the earliest wild card clinchers:
Yr Tm Lg Div ClinchDate LastTmElim Games Remaining W L PCT 2001 OAK AL W 20010923 MIN 12 102 60 .630 1997 NYA AL E 19970920 ANA 8 96 66 .593 1997 FLO NL E 19970923 LAN 5 92 70 .568 1999 BOS AL E 19990929 OAK 4 94 68 .580 2000 NYN NL E 20000927 LAN 4 94 68 .580 2004 BOS AL E 20040929 OAK 4 98 64 .605