About Me

My photo

Nice guy.  Have some blogs.  Do baseball research.

Monday, April 4, 2016

baseball-reference.com lists Russell freakin' Branyan tied with Mickey Mantle in AB/HR. Oh, the humanity.

This is sad. baseball-reference.com uses lazy old criteria for rate stats like batting average (BA).

While looking at AB/HR at baseball-reference.com I noticed Russell Branyan listed as number 16 among career leaders despite the fact that he hit only 194 home runs; Branyan qualified because he had at least 3,000 plate appearances (PA): 3,398. Of course PA is not used in AB/HR and it's a VERY low threshold. Branyan had 2,934 AB. Mantle had 8,102 AB and 536 HR. What the heck?


Statistic Description: At Bats Per Home Run
Minimum of 1000 IP, 3000 PA, 500 games (fielding, 500 IP for Ps), 200 stolen base attempts (catchers) or 80 stolen base attempts (baserunners only since 1951) or 100 decisions for career and active leaderboards for rate statistics.


baseball-reference.com lists Russell freakin' Branyan tied with Mickey Mantle in AB/HR. This could make your head explode.

Qualifying for Batting Average makes no sense. It doesn't use the denominator. Sunday, April 3, 2016

Denominator. You know. The bottom of an equation...

The same problem exists for slugging average. Both have at bats (AB) as the denominator, NOT plate appearances (PA). It's pretty basic.

There are other problems, which will be addressed in another post. They include whether the qualification should be based on opportunities (AB) or accomplishments (Hits, Total Bases, etc.) or some combination. However, opportunity qualification should NOT be based on PA.


1958 and 1980 AL examples: qualifier for BA lead explored. Monday, April 4, 2016

1958 AL Home Run King: Rocky Colavito or Mickey Mantle? Monday, April 4, 2016


What are the minimum requirements to lead a Rate Stat?
This is a bit of a dicey proposition, because the standards have changed quite a bit throughout time. Here is how we computed them for the website. Thanks to Bill Deane, Gerry Myerson and Total Baseball for clarifying some of these issues.
Batting Average, OBP, Slugging Percentage, OPS
  • Prior to 1920, a player must have appeared in 60% of the team's games to qualify for a title. This number was rounded to the nearest integer.
  • From 1920-1937 (unclear, and previously thought to be until 1944), a player must have appeared in 100 games.
  • From 1938-1944, the AL used 400 at bats and the NL stayed with 100 games, as discovered by Paul Rivard of SABR.
  • From 1945-1956, a player must have 2.6 at bats per team game. Note, however, that from 1951-1954 a player could lead if they still led after the necessary number of hitless at bats were added to their at bat total.
  • From 1957 to the present, a player must have 3.1 plate appearances per team game. Note, however, that from 1967 to the present a player could lead if they still led after the necessary number of hitless plate appearances were added to their at bat total.
In seasons where a player could still qualify for a title without the minimum plate appearances, we have printed out the altered entry with an asterik. For instance, in 1995 Mark McGwire didn't have enough PAs to qualify for the league lead in slugging, but when enough hitless at bats were added, so he qualified, his .636 (down from a real value of .685) still managed to place him fifth in the league. The real number is in his batting line and the altered number is in his leaderboard.

1. Who cares what Bozo criteria was used a million years ago?

2. Why would people as intelligent as those running baseball-reference.com use PA when it's not part of the stat? It's just going along with the establishment as baseball-reference.com did in 2012 when it pretended that Melky Cabrera did not have the highest BA because Melky had been suspended for using performance enhancing drugs (PED); Melky's agent had Melky send a self serving letter to feeble minded commissioner Allan Huber "Bud" Selig asking that he, Melky, not be honored with the batting championship, ignoring that it's a stat, not an honor.

I thought I had stumbled onto something cool that supplemented my recent research on HR Rate (AB/HR):


Year-by-Year Top-Tens Leaders & Records for AB per HR

Then I started to wonder what criteria had been applied, especially after I had addressed that very point in recent posts. Most of the leaders are probably legit but you've got to check since the site does not show either AB or HR. It's a bit cramped on this page but those basic stats are not listed for career leaders where there's plenty of room. Thank goodness for Russell Branyan sticking out like a sore thumb or I might not have checked.

Ah, that's better. Now my head won't explode.

No comments: