FREE hit counter and Internet traffic statistics from

Tuesday, February 06, 2007

Chat Transcript 02/06

The chat transcript from today can be found on the BP. Thanks to everyone who participated and all the great questions.

I was having some difficulty with the chat interface (special characters caused some posting problems) and so I had a couple of answers I wanted to get to but couldn't post. In any case here was one particular question I found interesting:

maxexpos (Montreal): Hey Dan, Could you rank this year's crop of Rookie...Managers? Thanks

Here is the entire answer....

Dan Fox: To recap, here are the five rookie managers for 2007 in no particular order.

  • Fredi Gonzalez – Marlins

  • Bud Black – Padres

  • Bob Geren – A's

  • Ron Washington – Rangers

  • Manny Acta – Nationals

  • All of these guys would seem to be players managers and certainly in the case of Geren, he better be since Ken Macha was fired for supposedly not being much of a communicator. There was a nice article in USA today on these guys and I especially liked this quote:

    And Gonzalez, 42, has been reading statistical books by baseball historian Bill James to get a better understanding of slumps. "Sometimes, you have to make a knee-jerk reaction, switching the lineup because a guy is 0-for-12," Gonzalez says. "The great players have a 0-for-20 or a 1-for-35. So, we've got to live with it."

    How can you not be rooting for Gonzalez after that?

    Washington will certainly be the most interesting to watch since he's kind of a character and is the ultra-player's manager it would seem from the comments I've heard him make this offseason. Acta and Gonzalez will probably get very long leashes since both teams are still developing. Black is also interesting in that he's a pitcher and is the only one to have never never managed in the minor leagues.

    Gonzalez, Washington, and Acta were all third base coaches recently and I've (with help from fellow SABR member Neal Williams) done a little work on seeing what effect those coaches have on overall advancement on hits (a derivative of EqHAR) in the period 2000-2006. Here are their number of opportunities (the number of opps that likely would have influence over), the number of runners who were thrown out, and the number of theoretical runs above what would have been expected given the quality and quantity of the opportunities and taking into account how the runners performed relative to other opportunities that the coach was not likely an influence on.

    Name Opp OA EqHAR
    Ron Washington 1730 45 10.6
    Manny Acta 1032 17 8.1
    Fredi Gonzalez 1249 25 -13.7

    Acta comes out on top on a per opportunity basis and ranks 29th of the 75 coaches that were looked at during the period. That said, I also discovered that there is no correlation from year to year in terms of coaches (both when they stay on the same team and when they move) and so there is probably very little to learn from this other than Acta was more fortunate and Gonzalez was not.


    David Gassko said...

    Hey Dan,

    Two comments about the chat:

    (1) It seems to me that busted hit-and-runs are a very serious issue in your base-running metric (I know you've acknowledged this; I'm just thinking out-loud here). I know they're not marked in PBP data, but perhaps you could get a decent proxy by looking at all stolen base attempts with less than two outs where the batter swings and misses. Perhaps you'd want to exclude 3-5 batters or something as well.

    It might not give you a perfect measure of failed hit-and-run plays, but it would be interesting to see the average eqSBR in that situation, and to see which players are most hurt there.

    (2) Someone asked about the predictive power of second-half statistics. I've actually looked into this for an article for Heater Magazine. Second-half numbers are slightly more predictive than first-half numbers, by about as much as you would expect given how we weight stats.

    If you think about it, when we project a player, we weight his last three seasons 5-4-3 or something like that, acknowledging that more recent stats are more predictive than older numbers. But you could take that further, and you would conclude that his second-half numbers should be weighted a little more than his first-half stats. In fact, we should be weighting each day (heck, each plate appearance!) a little more than the previous one.

    Of course, in terms of the actual projection, it will have little impact.

    Dan Agonistes said...

    Regarding point (1) I certainly agree with you. And yes, that's the approach I'm going to take in weeding them out. I'll then compare what I get there with the number of hit and runs tabulated in the Bill James Handbook and see where we are.

    As for (2) you are certainly correct as you reason. The example of a player I was going to mention in the chat but forgot was Dan Uggla who struggled a bit in the 2nd half, especially in September (.221/.269/.361)