Still consistent, still crappy.
I took a stab last week at comparing a team’s consistency in scoring runs to the style of ball they play. The result was a little counterintuitive- that last year, the Jays were more consistent than the best “small ball” teams, despite the idea that those teams can scratch across a run here or there when we need it, while the Jays sit back and wait for the home run.
I had a few requests to run the standard deviation for this season, so I did it for a few more teams this time:
I though the Jays would be higher this season because in May they hit .230 and slugged 40 home runs. But they’re pretty much where they were last year, which is at the lower end of the scale. And again, speed has no particular effect on the results.
I might be convinced into doing more teams if anyone can think of something (other than total runs scored) that might correlate with more consistent scoring, but for now I’m happy with the results:
- The Jays have not been an inconsistent team this year or last;
- Having a running game does not even out the runs you score.
The Mockingbird 
Umm… interesting but there’s a flaw in just running the standard deviation and making the claim that you are with this dataset. The flaw is this, you’re not taking into consideration the mean runs scored for each team which influences the meaning of the standard deviation in this case (the standard deviation of low scoring teams like the Jays will be smaller as there’s a natural “floor” in the data – a team can’t score less than zero runs).
Sure the Yankees have a larger standard deviation, but last I looked they score nearly a run/game more than the Jays (I just looked, NYY – 5.68 R/G, TOR 4.73 R/G)… to make a claim a like you are you’ve got take into account that the std is correlated with the mean level of run scoring. Not the best, but one way of correcting for the floor in the data set is diving the std by the R/G.
christopher taylor
July 27, 2007 at 9:51 pm
Ok- but I wasn’t really trying to rank dissimilar teams by consistency. If you look at the previous post it was concentrating on the difference between the Jays (as a “home run team”) and other speed teams that score the same numbers of runs. I think my two points still stand when you consider the Angels, Twins and Dodgers over the two years.
I threw in some of the best and the worst offences in baseball to try an give an idea what the range was – I guess I should have left in my paragraph about why the Yankees were so high (their number jumped half a point after their 2-day offensive explosion) and the Nats so low, but I figured that was being boring.
Is there a better easy/formulaic way to standardize the range than dividing by R/G? I wanted to compare standard deviation to whether or not teams underperform their projection, but gave up on the idea because of what you’re pointing out. I suspect even with a squashed range that comparing the Yankees and the Nationals is going to end up as meaningless as trying to compare the ‘consistency’ of David Ortiz and John McDonald…
halejon
July 28, 2007 at 3:44 pm