MBB: Michigan Intercollegiate Athletic Association

[Mr. Ypsi]

The topic of 'the hot hand' has been hashed out here, but something new from a different sport.  Dan Uggla has now set an Atlanta record hitting streak with 32 straight games - he is now hitting .224 for the season!  I would assume his average starting the streak must have been at or below .150.  I'm not quite sure how one would calculate the odds on this, but I would guess it must be somewhere north of a billion to one.

Despite definitely 'feeling' the 'hot hand', I agree with my fellow statisticians that it is a myth, but wonder how they will respond to this latest 'anomaly'.

Mr. Y,
Dan Uggla was batting .173 when he started his streak on July 5th. He is now up to .229 after his 3 for 3 game on Friday night. I don't know how hot his hands are but his bat is on fire! No one in the history of Major League Baseball has ever had a lower batting average with a streak this long. All year us Braves fans were waiting to see when he'd break out of his season long slump. I figured the Mendoza Line was going to become the Uggla Line. Glad to see him overcome all the adversity that came with the territory.   

If he breaks Joltin' Joe's record, he might reach .250!

[KnightSlappy]

The topic of 'the hot hand' has been hashed out here, but something new from a different sport.  Dan Uggla has now set an Atlanta record hitting streak with 32 straight games - he is now hitting .224 for the season!  I would assume his average starting the streak must have been at or below .150.  I'm not quite sure how one would calculate the odds on this, but I would guess it must be somewhere north of a billion to one.

Despite definitely 'feeling' the 'hot hand', I agree with my fellow statisticians that it is a myth, but wonder how they will respond to this latest 'anomaly'.

Uggla is a career .259 hitter (I'll round to .250), so he get's 1 hit every 4 at bats. For his career he's gotten 3.8 AB's per game (I'll round to 4).

The chance of not getting a hit in a game is about .75x.75x.75x.75 = .32. So chance of getting at least one hit in any one game = .68.

Chance of getting a hit in 33 straight = .68^33 = 0.0000036 (or 3.6 in a million).

[KnightSlappy]

Expanding on my numbers from yesterday, there were about 400 players that played in at least 33 games and gained enough plate appearances that they could have possibly put together a 33 game hit streak in 2010 (the average batting average of this group is about .270). Those 400 players give us about 32,000 unique 33 game stretches.

Just using these averaged numbers and doing some very rough calculations, we would expect to see a 33 game hit streak in baseball every 7 to 9 years.

[scottiedawg]

The topic of 'the hot hand' has been hashed out here, but something new from a different sport.  Dan Uggla has now set an Atlanta record hitting streak with 32 straight games - he is now hitting .224 for the season!  I would assume his average starting the streak must have been at or below .150.  I'm not quite sure how one would calculate the odds on this, but I would guess it must be somewhere north of a billion to one.

Despite definitely 'feeling' the 'hot hand', I agree with my fellow statisticians that it is a myth, but wonder how they will respond to this latest 'anomaly'.

Uggla is a career .259 hitter (I'll round to .250), so he get's 1 hit every 4 at bats. For his career he's gotten 3.8 AB's per game (I'll round to 4).

The chance of not getting a hit in a game is about .75x.75x.75x.75 = .32. So chance of getting at least one hit in any one game = .68.

Chance of getting a hit in 33 straight = .68^33 = 0.0000036 (or 3.6 in a million).

Your math is all right, but when players like Wily Taveras can have a 30-game hit streak I think we can all agree that there's a large amount of random variation that allows pretty unskilled players to sustain hit streaks.

Your math does illuminate the fact that DiMaggio's record may be the one in baseball with the smallest chance of ever being eclipsed.

[scottiedawg]

More stats! I'm sure this will excite a lot of people on this board.
http://blogs.prestosports.com/2011/08/20110811_free_stats

[Gregory Sager]

Your math does illuminate the fact that DiMaggio's record may be the one in baseball with the smallest chance of ever being eclipsed.

Nope. The baseball record with the smallest chance of ever being eclipsed is Cy Young's career wins record. Here's the all-time leaderboard in wins:

1. Cy Young	511	1890-1911
2. Walter Johnson	417	1907-27
3. Pete Alexander	373	1911-30
3. Christy Mathewson	373	1900-16
5. Pud Galvin	365	1875-92
6. Warren Spahn	363	1942-65
7. Kid Nichols	361	1890-1906
8. Greg Maddux	355	1986-2008
9. Roger Clemens	354	1984-2007
10.  Tim Keefe	342	1880-93
11.  Steve Carlton	329	1965-88
12.  John Clarkson	328	1882-94
13.  Eddie Plank	326	1901-17
14.  Nolan Ryan	324	1966-93
14.  Don Sutton	324	1966-88
16.  Phil Niekro	318	1964-87
17.  Gaylord Perry	314	1962-83
18.  Tom Seaver	311	1967-86
19.  Hoss Radbourn	309	1880-91
20.  Mickey Welch	307	1880-92
21.  Tom Glavine	305	1987-2008
22.  Randy Johnson	303	1988-2009
23.  Lefty Grove	300	1925-41
23.  Early Wynn	300	1939-63

... and here's the all-time leaderboard in hitting streaks:

1.  Joe DiMaggio	56	1941
2.  Willie Keeler	45	1896-97
3.  Pete Rose	44	1978
4.  Bill Dahlen	42	1894
5.  George Sisler	41	1922
6   Ty Cobb	40	1911
7.  Paul Molitor	39	1987
8.  Jimmy Rollins	38	2005–06
9.  Tommy Holmes	37	1945
10.  Gene DeMontreville	36	1896–97
11.  Fred Clarke	35	1895
11.  Ty Cobb	35	1917
11.  George Sisler	35	1924–25
11.  Luis Castillo	35	2002
11.  Chase Utley	35	2006
16.  George McQuinn	34	1938
16.  Dom DiMaggio	34	1949
16.  Benito Santiago	34	1987
19.  George Davis	33	1893
19.  Hal Chase	33	1907
19.  Rogers Hornsby	33	1922
19.  Heinie Manush	33	1933
19.  Dan Uggla	33	2011

The first thing you notice is that the gap between first and second place is roughly the same on both lists; both Walter Johnson and Willie Keeler have roughly 80% of the totals racked up by Cy Young and Joe DiMaggio. And from there the lists trail off with roughly proportionate amounts.

However, upon closer examination, these lists are very different. Note the dates. With the exception of the pitching-dominant 1950s and 1960s, just about every era of major league baseball is represented on the hitting-streak list. That's to be expected; the craft of hitting a baseball hasn't really changed much throughout the history of the game. But look at the all-time wins list. It's completely dominated by deadball-era pitchers. The deadball era (1876-1920) makes up only a third of major-league history -- and a significantly smaller percentage of pitching careers throughout major-league history, due to expansion -- but it produced a massively disproportionate number of all-time wins leaders. All five of the players atop the all-time wins leaderboard pitched in the deadball era for at least half of their careers, and 11 of the 24 pitchers who've notched 300 wins in their careers -- almost half of them, in other words -- were deadball-era hurlers.

The reason for this is because pitching radically changed from the deadball era to the liveball era. For one thing, doctoring the ball was outlawed in 1919 (although a few pitchers here and there, Gaylord Perry most notably, made a practice out of altering the ball long after it was forbidden to do so). Unable to make the ball move by putting foreign substances on it or cutting it, pitchers were forced to develop alternative methods of making the ball elusive, other than such standard pitches in the repertoire as the curve and the knuckleball. Thus, the slider was developed. But the problem with breaking pitches, particularly the slider, is that they add stress to the already unnaturally-stressful act of pitching, and thus throwing a lot of them often curtails career lengths. By contrast, you throw a spitball, a shineball, or an emeryball with the same standard motion as a fastball. Thus, added arm stress and the diminishing returns of a pitcher as he went deeper into the game without altered-ball pitches in his arsenal meant that the days of marathon pitching ended with the advent of the so-called "live" ball. In the nineteenth and early twentieth centuries, it was not uncommon for two or three pitchers to start most games; by the end of the deadball era, the four-man rotation was commonplace. Now, of course, major league teams rely upon five-man rotations, further cutting into the opportunities that a pitcher has to win ballgames. And then there's relief pitching. Where once a starter was expected to finish almost every game he started, no matter how badly he got knocked around, nowadays a pitcher very rarely finishes a game he starts, no matter how well he pitches. The increasing importance of the relief pitcher has further cut into the ability of starting pitchers to rack up wins, since a starter can only win a game if his team is ahead when he leaves the game (or moves ahead in the inning in which he is replaced by a pinch-hitter or pinch-runner), and once a reliever is in the game the outcome is thus out of the starter's control.

Lastly, take another look at those two lists. The first list is a Who's Who of Major-League Pitchers. Every one of the 24 pitchers on the all-time wins list is either in the Hall of Fame or will join it once he becomes eligible, with the possible exception of the steroids-tarred Roger Clemens. Now, look at the hitting streak list. It contains some great hitters -- Ty Cobb and George Sisler, two of the greatest hitters to ever pick up a bat, made the list twice apiece, and such immortals as Wee Willie Keeler, Fred Clarke, Pete Rose, Prince Hal Chase, and Rogers Hornsby -- to say nothing of the Yankee Clipper at the top of the list -- appear there as well. But the list also includes such dire mediocrities at the plate as Bill Dahlen (career BA .272), George McQuinn (career BA .276), Benito Santiago (career BA .263), and Dan Uggla (career BA .259). Several others ended their careers in the .290s or low .300s -- good numbers, but hardly Hall of Fame material in and of themselves.

In other words, what Uggla proved is that it's possible for just about anybody who swings a bat and wears a major-league uniform to go on a lengthy streak like this. It's much more statistically likely for a Sisler or a Cobb or a Rose to do so, as the Dean of Knight Slappy U. points out, but it's still possible for the Dan Ugglas of the world. That means that Joltin' Joe's streak is still theoretically within reach by everybody, but all the more so in particular by the great batting-average hitters of the day. Cy Young's record? It's completely out of reach, because the format of the game itself with regard to how pitching is handled, and the rules pertaining to what a pitcher can or cannot do to a baseball, has changed.

[scottiedawg]

You are correct that current pitcher usage makes that record unbeatable, so in a totally literal sense you are right. But considering records that are under no different circumstances than 50-100 years ago, DiMaggio's is tops.

[Mr. Ypsi]

Agree with the points by Greg and scottie, both, and will add a few that are also 'impossible' due to changes in the game: strikeouts in a season, 513 (Matthew Kilroy, 1886), wins in a season, 54 (Al Spalding, 1875), and complete games in a season, 75(!) (William White, 1879).  The strikeouts record is vaguely conceivable, but barring a drastic change in philosophies, no modern pitcher would ever be given the chance to even come vaguely close on the latter two.

[sflzman]

Well congrats to Uggla either way, on getting to 33. Though it's not a record, it's still an incredible feat.....

[KnightSlappy]

Randy Johnson retired in 2009 after 22 seasons with 303 wins and as one of the best pitchers of the last generation. He only made about 600 starts in his career, and didn't even record 500 decisions (wins and losses combined). That Cy Young guy is certainly out of reach

[Happy Calvin Guy]

Re: "unbreakable" records:

There is a fundamental difference between a "counting" stat (like career wins/hits/etc) and a "streak" stat (like a hitting streak), and the probability calculation for either one. 

It's kinda like comparing the relative magnitude of winning 2 D3 men's basketball national championships vs winning 32 consecutive MIAA games.   

[sflzman]

I still think the number 1 unbreakable record in baseball is the Ripken most consecutive games streak.

[Mr. Ypsi]

I still think the number 1 unbreakable record in baseball is the Ripken most consecutive games streak.

Maybe, but that's what everyone said about Gehrig's streak!

[almcguirejr]

I like the Tigers' pick up of Delmon Young today. He is only 25 years old.  Jackson, Boesch, and Young in the OF are all young along with Dirks.  Verlander, Porcello, Scherzer, and Pfister are all in their mid 20's.  Cabrera and Avila both under 30.  They have a good nucleus of talent for the next few years.
We have to find a shortstop and 2nd baseman and move Peralta to 3rd.


On another note;  Former Holland Christian football player, Nate Brink, was mentioned by Brady Hoke at Sunday's press conference.  Hoke mentioned him with about 4 or 5 other defenders.  Maybe Brink, a walk on, will see the field this year.  Hoke complimented him on his toughness.

[KnightSlappy]

I like the Tigers' pick up of Delmon Young today. He is only 25 years old.  Jackson, Boesch, and Young in the OF are all young along with Dirks.  Verlander, Porcello, Scherzer, and Pfister are all in their mid 20's.  Cabrera and Avila both under 30.  They have a good nucleus of talent for the next few years.
We have to find a shortstop and 2nd baseman and move Peralta to 3rd.


On another note;  Former Holland Christian football player, Nate Brink, was mentioned by Brady Hoke at Sunday's press conference.  Hoke mentioned him with about 4 or 5 other defenders.  Maybe Brink, a walk on, will see the field the field this year.  Hoke complimented him on his toughness.

I'm not sold on the need to move Johnny away from SS (good hitters can drive terriffic value at the position, and Jhonny' fielding has been 'fine' this year), but if they did, I'd rather see him move to 2B.