May 22, 2005

My sister Eileen sent me this article in which the Ten Greatest Individual Streaks in Sports are named, and she asked me what I thought of it. Bearing in mind that this sort of thing is always extremely subjective, I thought it was an okay list. If I were to quibble with anything, it's with the top placement of Joe DiMaggio's 56-game hitting streak. Not because it's not an impressive achievement, but because I think the odds are greater than you might think of it some day being bettered. Ichiro!, for example, by virtue of getting lots of at-bats (leading off and not walking), as well as hitting for a high average, may have as much as a 4% chance to tie Joltin' Joe if he hits like he did last year.

# 4.1: Percent chance of Ichiro Suzuki tying Joe DiMaggio’s 56-game hit streak, based on his 2004 statistics. Ichiro averaged 4.37 AB/G which gives him an 86.9% chance of getting a hit in a game (1-(1-.372)^4.37). Based on that percentage, Ichiro has a 0.039% chance of hitting in 56-games in a row. Assuming he plays 161 games again, that’s essentially 105 consecutive attempts at the record. 0.039%^105 is 4.1%. However, this doesn’t take into account the fact that with every hitless game, Ichiro drastically reduces the number of chances he has for a 56-game streak, so his odds are significantly lower.

# 0.3: Percent chance of Ichiro Suzuki tying Joe DiMaggio’s 56-game hit streak, based on his 2004 statistics. No, this is not a misprint. An alternative calculation, using the fact that Ichiro hit successfully in 134 out of 161 games (83.2% instead of the 86.9% above), reduces his chances by over 91%.

I didn't say the odds were good, just not impossible. They're lower now that the season is almost two months old, but there's always 2006 and beyond. And there's no law that says there can't be another Ichiro!-like player on the scene some day.

One streak that never gets mentioned, but which in my opinion is a true Record That Will Never Be Broken, is Johnny Vander Meer's streak of two consecutive no-hitters. Two in a row isn't really thought of as a streak, but since it'll take three straight no-nos to beat him, I'll put that up there as a feat for the ages. It's rare enough for a pitcher to throw two no-hitters in a career, rarer still for them to come in the same season. A hundred years from now, I expect Vander Meer to still be unique.

Beyond that, not much to add. It's certainly a reasonable enough list. What do you think?

UPDATE: Bogey comments.

Posted by Charles Kuffner on May 22, 2005 to Other sports | TrackBack

I haven't thought of Vandermeer's feat in a while, but I'd agree that it's virtually impossible to replicate, much less exceed.

The most amazing thing about DiMaggio's streak in my mind is that had it not been for two great plays by Ken Keltner, the streak could have been 83; DiMaggio was shut out that game but then hit in the next 17. (My numbers may be off by one or two, but "you can look it up," to quote Casey.)

Posted by: Linkmeister on May 22, 2005 1:31 PM

Oops, 73. My math skills are weak this morning.

Posted by: Linkmeister on May 22, 2005 1:32 PM

WTF happened to Lance Armstrong's 6 consecutive Tour de France wins?

Posted by: JR on May 22, 2005 8:01 PM

The authors don't think too much of Cal Ripken's 2632 and Lance Armstrong's 6-peat. Those are serious athletic accomplishments that far exceed DiMaggio in terms of athletic accomplishment. I'm not trying to take away from DiMaggio, but the 56 games has a far higher element of luck and fortune than the grind that Ripken and Armstrong went through. Same thing with Edwin Moses. He went out and beat all-comers for 9 years. That's something.

Perhaps a more valid discussion would be separate the freaky "beat-the-odds" sort of streaks from the serious lifetime athletic accomplishments.

Oh, and I have no doubt that someone will come along and crush DiMaggio's record just like they did to Ruth and Maris. Surely there are more pure scatterball hitters like Ichiro out there and one will come along and put the ball in play for more than 56 games.

Posted by: kent on May 22, 2005 8:35 PM

Everyone talks about Gretzky's 51-game point scoring streak, but I think the fact that he had 13 consecutive seasons scoring 100 or more points is more impressive. Especially considering the closest anyone else has come is six.

Posted by: Pete on May 22, 2005 10:29 PM

None of these streaks hold a candle to the streak of the "Dipsea Demon", Jack Kirk, who ran the difficult Dipsea Race 68 straight years, with his last one in 2003 AT AGE 96, only ending the following year because he had fallen and broken his hip shortly before the race. And for many of those years, Kirk was competitive in the race, winning it two times.

Posted by: Steve Patt on May 23, 2005 12:05 AM

The Unitas streak is much more impressive than DiMaggio's. The reason nobody knows about it is that nobody has even remotely come close to it. Considering the era he played in with much less passing, it is truly an amazing accomplishment.

The Armstrong accomplishment is good, but Indurain won five in a row and Armstrong spends his entire year preparing for those three weeks in France. He's not racing on many other tours.

Ripken?? When your "accomplishment" actually hurts your team's performance it isn't something to boast about. Nobody will ever break that streak because nobody will ever be that selfish in the game of baseball again.

Posted by: Double B on May 23, 2005 1:28 AM

Having looked at all of those streaks, to me the most impressive is Moses. His streak involved no margin for error. DiMaggio had days where he went 1 for 4. Unitas' streak could include games where he threw 1 TD and 3 INTs and his team loses by 20. Moses won every race at that distance for over 9 years against the best in the world.

And BTW, Wilt's record is nothing to sneeze at either. Not only does it mean he had the endurance to last for 45 games but the discipline not to foul out.

Posted by: Patrick on May 23, 2005 8:04 AM

Indurain was a great rider but he's one of four who won five Tours. With 6, Armstrong stands alone.

Posted by: JR on May 23, 2005 11:16 AM

I think you have something Patrick. Athletics is about winning, not some statistic that is only peripheral to winning the event. That's why I think Armstrong and Moses have two of the most impressive accomplishments.

I also noticed that this article only discusses individual streaks, not team accomplishments. How about UCLA winning 10 NCAA basketball championships in 12 years including 7 in a row from 1967-1973? It's going to be a LONG LONG LONG time until any team ever matches that accomplishment in any major college sport much less basketball.

Posted by: Kent on May 23, 2005 11:20 AM

I'm not familiar enough with baseball statistics to be sure, but the first calculation seems wrong. As I understand it, batting average doesn't count walks, nor do walks count for DiMaggio's streak - it has to be a hit in each of 56 consecutive games. So unless average at-bats/game also doesn't count walks, the first calculation overestimates Ichiro's odds, since the chance of not getting a hit during a single at-bat is greater than 1-.372.

The second calculation is closer, but even it omits the fact that the odds of each possible 56-game streak in 161 consecutive games are not independent. If you miss in the 56th game, the odds drop to zero not only for games 1-56, but also for games 2-57, 3-58, and so on up to games 56-111. Thus the odds are even less than 0.3%.

I'll get back to you when I have the correct calculation.

Posted by: Mathwiz on May 24, 2005 5:23 PM

Mathwiz - That calculation most definitely excluded walks. "At-bats" strictly refers to times at bat that end in hits, outs (excluding sacrifices and sac flies) or errors - walks and hit-by-pitches are excluded. The term that encompasses everything is "plate appearances". So that part of the calculation is correct.

Posted by: Charles Kuffner on May 24, 2005 5:57 PM

Thanks for the clarification, Kuff. Of course, that still leaves this problem in both calculations:

However, this doesn’t take into account the fact that with every hitless game, Ichiro drastically reduces the number of chances he has for a 56-game streak, so his odds are significantly lower.

I tried to come up with a formula for the exact odds, but the problem turned out to be too hard for me. (It turns out to be similar to a binomial function, but with modified, tough-to-compute coefficients.)

I thought about writing a computer program to do a Monte Carlo simulation, but decided instead to just guesstimate, based on this idea: the number of attempts he has at the record isn't 105 (161-56) but about 14 (once at the beginning of the season, plus the number of times he'd be expected to blow the previous attempt with at least 56 games remaining in his season, so he can start another attempt. That's still not exact, but it should be closer). Using that idea with the numbers in the first calculation gives odds of a bit over 0.5%, or 1 in 186, rather than the 4% they came up with.

Which means your overall conclusion is still correct: the odds aren't very good, but they're not impossible either. As long as there's at least one player about as good as Ichiro in major league baseball every season, the odds of seeing DiMaggio's record tied or broken during our remaining lifetimes aren't as remote as the late Stephen J. Gould believed.

Posted by: Mathwiz on May 27, 2005 10:13 AM

I do not know if this counts for anything but on Sept 22 2005 I played in my 500th consecutive game in a 40& older hockey league. (1988 to present date) I am 70 and am still going strong. I cannot find a individual sports streak that has gone on for 17 years. Bruno Colapietro, Endicott NY

Posted by: Bruno Colapietro on September 25, 2005 4:18 PM