Comments on: What Is Luck?

By: Geoff Young

Geoff Young — Thu, 28 Jun 2007 13:32:36 +0000

#34: Thanks, Jay, for the thoughtful response. This is a difficult topic, and I'm not sure we'll ever reach a satisfying resolution, but it's good to think about these things every once in a while.

Current score: 0 (to vote for this comment, please visit the site)

By: jay

jay — Thu, 28 Jun 2007 07:00:30 +0000

A little late in this conversation, but I think of luck as the distance from the expected outcome vs. what actually occurs, especially the clustering of infrequent events. I don’t have any way of measuring it, but I am guessing if you ran a perfectly controlled experiment of a Scott Linebrink vs. a Fransden, ran the outcomes a billion times, you are going to get the “true” expected outcome of that match up. In some of those tests, Fransden hits a home run (not often), in some of those tests, Fransden K’s (a lot more). Now Fransden’s dribbler was not only a hit against Linebrink but a weak hit, but I think of it in the “hit” category. I think the “true” outcome of Linebrink vs. Fransden favors the Padres, so when Fransden wins, it is at odds with the expected outcome, but not incredibly shocking either (Linebrink is not that much better than Fransden), but still “lucky” for the Giants.

I am starting to stretch it here, but I would argue it would not necessarily appear like luck if Linebrink pitched a fat fastball that went for a jack. No one would say that Linebrink was “unlucky”, just bad. But, unless pitching fat fastballs was part of new trend, I would say that event, because it happens infrequently, is bad luck.

With the clustering events, I think of a team is given an allotment of 6 singles (no walks) in a game. If those are spread out over the game, the team will most likely get shut out. They are bunched sequentially, six in a row, that could generate 3 or 4 runs. To me, that is an improbable, aka “lucky”, outcome. So, I would use stats to predict the expected runs scored per 9 for six singles without a walk. If the team scores worse, unlucky, if the team scores better, lucky.

So when James says “1 run games are luck”, I interpret him to say the variability in outcomes, even between two mismatched teams but ones locked in a close scoring game, is such that which team prevails is a matter of chance, or luck. You play those two teams against each other a billion times, the better team will win more often, but when you isolate to the games where it is very close, a lot of variability/luck comes into play. I would expect a slight bias towards good teams winning those more often than weaker teams, just because the cards are stacked in their favor, but it is far from a lock.

I’ve re-read this a few times, and while it mostly makes sense to me, I expect it will muddy the discussion instead of clarifying it, but I thought I would add my thoughts.

Current score: 0 (to vote for this comment, please visit the site)

By: SDSUBASEBALL

SDSUBASEBALL — Wed, 27 Jun 2007 20:00:07 +0000

31: How about none of them. I dont really like any of those names.

Current score: 0 (to vote for this comment, please visit the site)

By: Didi

Didi — Wed, 27 Jun 2007 19:43:35 +0000

30: Randy Johnson throwing the slider and hitting the bird flying through the strike zones, thus decimating the bird into puffing feathers was a CHANCE.

Current score: 0 (to vote for this comment, please visit the site)

By: Didi

Didi — Wed, 27 Jun 2007 19:42:17 +0000

I wonder what the Cubs are going to do with their glut of OFs.
They don’t seem to know what to make of Felix Pie, Matt Murton, and Angel Pagan. Which one should the Padres try to trade for and at what cost?

I’m not sure that the Padres even have enough chips to trade for any of them.

Current score: 0 (to vote for this comment, please visit the site)

By: PF4L

PF4L — Wed, 27 Jun 2007 19:38:20 +0000

Re 12:

wiktionary definition of luck

http://en.wiktionary.org/wiki/luck

1. Something that happens to someone by chance, a chance occurrence.
2. A superstitious feeling that brings fortune or success.

I don’t believe in superstition, so I go with def #1.

As for def # 1, I don’t think it applies since ballplayers are technically completely in charge of what they do. A better hitter has more control over where he hits the ball (i.e. not directly at a fielder) and thence lowers his “chances” of getting out. But, he is definitely in control, as are the pitchers, fielders, coaches, umps, etc. Since no one is perfect, there are a lot of missed opportunities, but I don’t ascribe 99.9% of missed opportunities to “chance”.

Now, when Dave Winfield hit the bird with a thrown ball and killed it. THAT was chance.

Current score: 0 (to vote for this comment, please visit the site)

By: SDSUBASEBALL

SDSUBASEBALL — Wed, 27 Jun 2007 19:31:06 +0000

12-15 is a statistic that doesnt say we have the strongest bullpen in MLB.

Current score: 0 (to vote for this comment, please visit the site)

By: PF4L

PF4L — Wed, 27 Jun 2007 19:27:55 +0000

“pythagoreaically”…….nice!

Current score: 0 (to vote for this comment, please visit the site)

By: Eric

Eric — Wed, 27 Jun 2007 19:04:57 +0000

our bullpen is not really top-heavy, so we're not particularly "lucky," pythagoreaically speaking

Current score: 0 (to vote for this comment, please visit the site)

By: Eric

Eric — Wed, 27 Jun 2007 19:03:51 +0000

re: correlation between performance in close games and bullpen strength

I always thought it was too simple to say it’s about a strong bullpen…I think top-heavy bullpen makes more sense. If there was a great disparity between your best and worst relievers, think about it:

-your small leads will often be preserved by your hi-leverage relievers

-your large leads will often turn into small leads as your crappy mop-up men hand the game over to your hi-leverage guys

-your deficits will often turn into bigger deficits

hence you’d see a lot of close wins and blowout losses, and a “lucky” pythagorean split.

Current score: 0 (to vote for this comment, please visit the site)