strange luck

i'm as surprised as anyone could be, i think, to see the cubs come out of the weekend having won 13 of 17 and closed the gap between themselves and the division-leading milwaukee brewers by 5 full games in just 17 played. i've found myself laughing in disbelief more than once since that june 23 launching point. the cubs have since been one of the hottest clubs in baseball. the day before it began, i found reason for further pessimism looking forward. subsequently i further said:

should the starting pitching finally suffer for that eventual mean reversion while the offense -- now without barrett for the duration and playing the likes of rob bowen and koyie hill, as well as mike fontenot, ryan theriot, felix pie, angel pagan and jacque jones more days than not -- continues to struggle, the cubs may yet have harder days in front of them than any we've so far seen.

well, the pitching in this run has posted a team .310 babip over these 17 games -- and managed to hold the opposition to 4.24 runs/game anyway. the offense has moreover, for all the dead weight it may carry, posted 5.35 runs/game in the same time. goes to show you what i know about the future.

going 13-4, though, usually isn't enough to close five games on anyone unless they're not playing particularly well. and the brewers actually haven't played poorly at all -- since june 23, they've scored 87 and allowed 76. that hasn't prevented them from winning just 8 of 17, though, to the cubs' good fortune. all thoughts of selling into the trade deadline are likely now reversed.

ccd's advice is nonetheless well taken, because it still really is hard to make up five games. that's a point that probably convokes laughter in some quarters right now because it's altogether too easy to believe that what has just happened somehow had to, or that it was somehow a common enough occurrence that no one should be surprised when it happens.

a look at some data helps dispel the notion. over the last ten seasons (from the beginning of 1998), for just 31 of 1379 sample periods have the cubs made up as many as 5 games on the division leader or the next-best team in the division. that's 2.2% of the time. these constitute six separate runs against the leader, broken down as follows.

Gm	Date			Opp		RS	RA	YTD		Rank	GB	against the leader over last 17	last 17
								W	L				W	L
88	Jul	13	2007	HOU	W	6	0	45	43	2	3.5	5.0	13	4
89	Jul	14	2007	HOU	W	9	3	46	43	2	3.5	5.0	13	4
90	Jul	15	2007	HOU	W	7	6	47	43	2	3.5	5.0	13	4
113	Aug	9	2006	@MIL	L	3	6	48	65	5	13.0	5.0	11	6
114	Aug	10	2006	@MIL	L	6	8	48	66	5	14.0	5.0	11	6
117	Aug	13	2006	@COL	W	8	7	49	68	5	13.0	6.0	10	7
118	Aug	14	2006	@HOU	W	3	0	50	68	5	12.5	5.5	10	7
23	Apr	25	2003	@COL	W	11	7	14	9	1	-3.0	5.0	11	6
120	Aug	14	2003	HOU	W	7	1	63	57	3	0.5	5.0	12	5
121	Aug	15	2003	LAD	W	2	1	64	57	1	-0.5	5.0	12	5
56	Jun	6	2001	STL	W	4	1	35	21	1	-4.0	6.5	14	3
57	Jun	7	2001	STL	W	4	3	36	21	1	-5.0	8.5	15	2
58	Jun	8	2001	@CHW	L	3	7	36	22	1	-5.0	8.5	15	2
59	Jun	9	2001	@CHW	W	4	3	37	22	1	-5.0	7.5	15	2
60	Jun	10	2001	@CHW	L	1	3	37	23	1	-5.0	7.5	14	3
61	Jun	12	2001	@ARI	W	6	2	38	23	1	-6.0	7.5	14	3
62	Jun	13	2001	@ARI	L	3	13	38	24	1	-6.0	6.5	13	4
63	Jun	14	2001	@ARI	L	2	3	38	25	1	-6.0	6.5	12	5
64	Jun	15	2001	MIN	W	5	3	39	25	1	-6.0	6.5	12	5
65	Jun	16	2001	MIN	W	11	4	40	25	1	-6.0	6.0	12	5
66	Jun	17	2001	MIN	W	5	4	41	25	1	-6.0	6.0	12	5
100	Jul	27	2000	@PHI	W	4	1	46	54	3	10.5	6.5	13	4
101	Jul	28	2000	SFG	L	0	2	46	55	3	10.5	6.0	13	4
102	Jul	29	2000	SFG	W	8	1	47	55	3	9.5	6.0	13	4
103	Jul	30	2000	SFG	W	3	1	48	55	3	8.5	7.0	13	4
104	Jul	31	2000	COL	W	2	0	49	55	3	8.5	8.0	14	3
105	Aug	1	2000	COL	L	1	2	49	56	3	8.5	8.0	13	4
106	Aug	2	2000	COL	W	3	2	50	56	3	8.5	7.0	13	4
107	Aug	3	2000	@SDP	L	5	6	50	57	3	9.0	5.5	12	5
108	Aug	4	2000	@SDP	L	9	11	50	58	3	9.0	5.5	11	6
109	Aug	5	2000	@SDP	W	6	3	51	58	3	9.0	5.5	12	5

this is clearly not a common event. as an aside, winning 13 of 17 is rarer still -- it's been true of just 24 of the 1379 sample periods.

it's certainly fair to note that -- with six discrete events in ten seasons -- at some point the cubs are likely to do such a thing in most years, and one can sit in wait of a charge of this kind when the club's hovering within single digits of the lead. but that ignores the probability of the opposite dynamic -- it's actually been more than four times as likely that the club finds a way to drop five or more against the lead in the same time. by my count (and it's a bit subjective, but my arbitrary view is that any stretch without a five-game break constitutes a continuing streak) there have been 26 such episodes, including one earlier this season. and it's not a phenomena of abject losers alone -- 13 of the 26 involved clubs that were placed third or higher in the division at some point in the run (usually at the start).

this is something it might pay to note if you're currently suffused in the headiness of winning baseball. by all means, enjoy it -- i certainly am -- but consider the history before you bet your friend from milwaukee a hard-earned c-note straight up that the cubs will take the division this year. get yourself some odds on that one -- you deserve at least 3:1 and probably more. not only is there a non-negligible chance that a five-game retracement lies somewhere in the future of this very same cubs club; but you'd effectively be betting that the lightning we've witnessed in the last three weeks is going to strike twice this year after hitting just five times in the previous nine seasons.