Rob Neyer and fact-checking Robothal

by Adam J. Morris on Aug 26, 2008 12:45 PM CDT Comment 10 comments

Jeff (Boston, MA): I heard Ken Rosenthal last night say that the Minnesota Twins are an exception to the all the statistics on a year in year out basis and that their high batting average with RISP is skill and NOT a fluke. I call shenanigans although I do think Ken is a very good writer.

Rob Neyer: This is an empirical question, right? I'm not going to call out Robothal without checking ... But if someone wants to look it up, I'll be happy to present the information and let you be the judges.

Thanks to B-R, this is easy enough to check out...

Year	Minn Avg. Overall	Minn. Avg. RISP	AL Avg. Overall	A.L. Avg. RISP
2008	.279	.313	.267	.272
2007	.264	.276	.270	.275
2006	.287	.296	.275	.276
2005	.259	.271	.268	.273
2004	.266	.277	.270	.272
2003	.277	.268	.267	.273
2002	.272	.269	.264	.271
2001	.272	.252	.267	.268

Okay...so for the A.L. as a whole, average with RISP over this period is generally 2-5 points higher than overall average.

And from 2001-03, the Twins consistently hit well worse with RISP than overall.

But starting in 2004, the Twins have consistently hit better with RISP than overall -- anywhere from 9-12 points better before this season, when they are hitting 34 points better with RISP.

I'm not sure that's really statistically significant. However, if one wanted to hit for a higher average, one could conceivably achieve that by swinging at more first pitches...this year, the league as a whole is hitting .338 when putting first pitches into play. The downside of that, of course, is that you also have a .338 OBP, and if you swing and miss or foul the pitch off, you are in an 0-1 hole, and your spread between average and OBP would go down.

Year	Minn spread Overall	Minn. spread RISP	AL spread Overall	A.L. spread RISP
2008	59	76	68	85
2007	66	71	67	81
2006	60	77	64	81
2005	64	83	62	75
2004	66	79	68	83
2003	64	73	66	81
2002	60	66	67	81
2001	65	88	67	83

So...the spreads, overall, during this span are from 13-17 points. The Twins' spreads are, for the most part, in that range, except in 2002, 2003, and 2007, when the spread drops into the single digits. Which doesn't help, because in 2002 and 2003, when the spreads were smaller (and one would, by the "swing early" theory, expect higher averages with RISP, the Twins had a lower average with RISP).

So I don't have a good explanation for what is happening with Minnesota with RISP this year, or over the past few years. I'm inclined to chalk it up as a statistical fluke, though.

0 recs | Comment 10 comments

Story-email Email Printer Print

Comments

Is Jeff (Boston, MA)

our buddy from Lookout Landing?

Offense doesn't doubt me, but my first and primemost thing is defense and punt return and kickoff return

by Brett Perryman on Aug 26, 2008 12:49 PM CDT reply actions 0 recs

I doubt it

I think he’s moved back to San Diego.

by Adam J. Morris on Aug 26, 2008 12:52 PM CDT up reply actions 0 recs

Sullivan probably wouldn't

commend Rosenthal at the end like that either.

Offense doesn't doubt me, but my first and primemost thing is defense and punt return and kickoff return

by Brett Perryman on Aug 26, 2008 1:23 PM CDT up reply actions 0 recs

Whether or not

you chalk it up to statistical noise, it begs that Rosenthal be asked: what change did the Twins make in 2004 to “make” this happen?

Time you enjoy wasting was not wasted.

by t ball on Aug 26, 2008 1:28 PM CDT reply actions 0 recs

nah

i would suspect low strike out totals would explain it.

by SteveP on Aug 26, 2008 1:51 PM CDT reply actions 0 recs

I'd think

Without evidence, that the quality of pitching faced would make at least a situtational difference in the case of the Twins, and possibly any other anomaly in "statistical norm’ having to do with RISP. But that would be a lengthy, detailed, intense study which still might not be meaningful.

"He uses statistics as a drunken man uses lamp-posts... for support rather than illumination." - Andrew Lang (1844-1912)

by Ed Coffin on Aug 26, 2008 2:06 PM CDT reply actions 0 recs

The way to figure this out

would be to look at the distribution across all teams. It may be that most teams are usually within 10 points of the league’s difference, so the Twins 34 points may be quite significant. But I’m too lazy to dig through it that much.

Without looking at the spread of different teams, then assuming no trend, you’d expect any given team to finish with a BAARISP-BAA greater than league average half the time.

1 year better than average: 50% — 7 AL teams
2 years better than average: 25% — 3.5 AL teams
3 years better than average: 12.5% — 1.75 AL teams
4 years better than average: 6.25% — 0.875 AL teams
5 years better than average: 3.125% — 0.4375 AL teams

This suggests that about once every 2 five-year periods (I guess these can be partially overlapping) you will have around one team that is on a 5 year trend like the Twins.

This is the null hypothesis, and basically suggests that if everything were completely random, you’d see a team on such a run about half the time (again, this is non-parametric and disregards their magnitude).

by JBImaknee on Aug 26, 2008 4:03 PM CDT reply actions 0 recs

Another problem

If hitting with runners in scoring position were indeed a skill, how could it have transferred from the 2004 team to the 2008 team? Those two teams share exactly zero starting position players in common. The 2005 and 2008 teams share exactly three starters in common.

If you want to make the argument — as Adam has suggested — that the Twins may have a particular organizational approach to hitting that leads to higher RISP (e.g. by swinging more often at first pitches), then that’s fine. Whether or not that’s actually the case is to be debated, but regardless, conflating an organizational mandate with a collective (or individual) skill is tomfoolery.

by Eric Simon on Aug 26, 2008 5:13 PM CDT reply actions 0 recs