Two Kinds of Luck

It’s that time of year when many GMs are deciding between acquiring pieces for the stretch run and calling it a year and selling valuable spare parts. An important part of the equation is determining how well the GM thinks his team will perform for the rest of the year. In general, teams with better records will be more inclined to buy than sell, and vice versa.

But, a raw win-loss record can be misleading. In some cases, the team that is scoring and allowing runs that would normally result in more or fewer wins may indicate that the team is better or worse than its actual record. Let’s call this Pythagorean luck (I dislike the confusing reference to the Greek mathematician, but the name has stuck). For example, the Braves this year are an unimpressive 47-52; however, the Braves have played much better than their record. They have scored 440 runs while allowing 412 runs: a combination that ought to produce a winning record of approximately 52-47.

On the hitting side, the Braves have been slightly below league average, scoring 4.44 runs per game, compared to the NL average of 4.51. It’s the pitching that has really kept the Braves playoff hopes alive. The team’s 3.81 ERA is second in the league. However, this is where I want to bring in the other type of luck a GM needs to consider.

Frank Wren, or any other GM, cannot just look at a Pythagorean W-L record and assume random run-agglomeration will even out and expect that as luck turns around that his team will be good enough to contend. The other type of luck is the random fluctuation of player performances around their true talent level—let’s call it talent luck. In the case of the Braves, I believe the pitching has performed better than its talent level and will not hold up. Hats off to Roger McDowell and the pitching staff, but I don’t think the team will finish second in ERA. Now, it is possible that the pitching will continue to be excellent as it has been, but I doubt it.

And even if the pitching and hitting continue on their current trajectories, the team still faces a significant real six-game deficit in wins to a Philadelphia team that has also played a bit unlucky: 53-46 compared to a 57-42 expected record.

I wonder if this is similar to the conclusion that A’s GM Billy Bean reached about his Oakland team. Though the A’s have outperformed their record by four wins, and the division-leading Angels have won 7 games more than the run differential predicts. Yet, Beane traded away both Rich Harden and Joe Blanton to the Cubs and Phillies, respectively.

7 Responses “Two Kinds of Luck”

  1. Rick says:

    Two good trades by Beane. Especially with Harden. He’s going from one of the most extreme pitcher’s parks to one of the most extreme hitter’s parks. Granted, pitchers from the AL tend to do better in the NL, but the park factor plus his injury history could really come back to bite the Cubs on the ass.

  2. Dayan says:

    I heard that there was a theory by James that said that to improve your team in the second half you must improve the area in which you are better. So if your pitching is strong, adding a good arm is going to give you more wins than adding a good bat to your “weak” ofense.

    What do you know about this, JC?

  3. JC says:

    I’m not familiar with the theory. My feeling is that runs are runs. You can prevent them or score them. If your pitching is good, but you can’t get another bat, then there is nothing wrong with getting more pitching.

  4. Dayan says:

    Sounds logic to me that you can outperform your Pyth record with good bullpen, as they play a large role in close games. Do you think we can find significance to that?

  5. Cliff says:


    That, bulpen performance, is the one area that seems to be a proveable and consistent separator of “Pythgagorean” records.

    And actually, runs saved are slightly more valuable than runs scored. That is, it is impossible to lose when the other guy scores 0. However, even if you score 18, the other guy can score 19 (like the Tigers last night). Also, when the run “market” is lower (60’s, say), then each run of differential is worth more.

  6. dan says:

    I like the new look, but the font is absolutely tiny. I’m 17 with 20/20 vision and am having difficulty reading this quickly.

  7. Edward says:

    Dr. Bradbury,

    How does Joe Blanton compare to other pitchers whom you thought were under-valued by other teams, i.e. Aaron Harang? As long as Blanton doesn’t eat as many cheesesteaks as he does innings, I’m happy with the trade.