I told him that a player on a streak has to respect the streak.
— Crash Davis in "Bull Durham"

Jimmy Rollins just ended his 38-game hitting streak and, after one hitless game, started another streak that now stands at four games after his 3-for-4 showing last night. It was the second longest streak since Pete Rose's historic 44-game hitting streak in 1978 as well as the second longest multi-year streak (after Wee Willie Keeler's 45-gamer that started in the last game of the 1896 season).

Oddly, when I tried to look up the complete list of 30-game hitting streaks, I got wildly divergent lists. I cobbled together what I believe is a complete list from three main sources (1, 2, and 3):

56—Joe DiMaggio, 1941
52—Denny Lyons, 1887 (inc. walks as hits)
45—Willie Keeler, 1896-97 (last game of 1896 plus first 44 of 1897)
44—Pete Rose, 1978
42—Bill Dahlen, 1894
41—George Sisler, 1922
40—Ty Cobb, 1911
39—Paul Molitor, 1987
38—Jimmy Rollins, 2005-06
37—Tommy Holmes, 1945
36—Billy Hamilton, 1894
35—Luis Castillo, 2002
35—Ty Cobb, 1917
35—Fred Clarke, 1895
34—Benito Santiago, 1987
34—Dom DiMaggio, 1949
34—George McQuinn, 1938
34—George Sisler, 1925
33—George Davis, 1893
33—Heinie Manush, 1933
33—Hal Chase, 1907
33—Rogers Hornsby, 1922
31—Vladimir Guerrero, 1999
31—Ken Landreaux, 1980
31—Rico Carty, 1970
31—Willie Davis, 1969
31—Sam Rice, 1924
31—Nap Lajoie, 1906
31—Ed Delahanty, 1899
30—Albert Pujols, 2003
30—Luis Gonzalez, 1999
30—Eric Davis, 1998
30—Sandy Alomar Jr., 1997
30—Nomar Garciaparra, 1997
30—Jerome Walton, 1989
30—George Brett, 1980
30—Ron LeFlore, 1976
30—Stan Musial, 1950
30—Goose Goslin, 1934
30—Tris Speaker, 1912
30—Elmer Smith, 1898
30—Cal McVey, 1876

The one entry on the list that you never see is Denny Lyons' 52-game streak in 1887. That year the National League counted walks as hits. Lyons kept the streak going a few times without registering what we consider a hit today. MLB decided to reset the stats from 1887 to reflect our current definition for hits. However, the fact remains that Lyons had a 52-game streak according to the rules in play at the time.

You'll note that Rollins comes in ninth in terms of longest hitting streaks all time.

I thought it might be interesting to look at the player stats for all of those registering a thirty game streak or longer. Given their stats we can then determine the odds that each player would have hit the streak that he did.

I established the statistics involved in deriving the odds in a post from a few years ago. I have changed it slightly to ignore those plate appearances in which the batter had no chance to get a hit (i.e., times hit by a pitch and intentional walks).

For each, I list the odds that the batter will get a hit in any given plate appearance, in any given game, for a stretch of games as long as their streak, and for a stretch as long as their streak over their season(s):

So basically, the most likely was Big Ed's 31-game streak in 1899, and the least likely was DiMaggio's 56-gamer.

But isn't that a function of the length of the streak. It's just harder to get a hit for 25 more games, right? Well, I put that theory to the test. I projected each hitter's odds of recording a 56-game streak. Who had the best chance?:

So DiMaggio was just the twentieth most likely to record a 56-game streak. That's about half way down the list and miles away from Lyons' 43% likelihood.

That made me wonder who in baseball history had the best shot to record a 56-game hitting streak in any single season. I ran the numbers and—surprise!—most of the players are from 1887 when walks were counted as hits. For each, I translated the 56-game probability to its "one in X" equivalent:

Yeah, that Tip O'Neill was a shirker. He should have had at least a couple of 56-game streaks.

I started the data at 1900 to filter out those 1887 anomalies:

And just to put it in perspective, I looked at the most likely to get a 56-game hitting streak since 1941. Here's the list up to DiMaggio:

So what does this tell us? Well, mainly it says that DiMaggio's streak may have been one of the least likely events to ever occur in the game. But I think it says that DiMaggio's streak is not unattainable. Ichiro two seasons ago had the best shot since 1941 to break it. But as we saw when we compared streaks of various length, extending a long streak to an historic one is a difficult or at least unlikely thing. And it doesn't get any easier when the media start hounding you to answer whether you'll get a hit the next day while you have a towel wrapped around your midsection.

That said, I wouldn't be surprised to see someone top the historic 56-game figure. I say we go back to counting walks as hits to move the process along.