19 Responses to “So, You Had A Bad Day…Or Two”

1. zuty Says:
   December 29th, 2009 at 11:10 am

   As a curiosity, I looked at one of the box scores from one of these games -- Pete Rose's game on July 19, 1966. My curiosity was further piqued when I saw that he scored a run in that game, which led me to wonder: How can a player who reaches base zero times in a game score a run?

   Turns out he reached on a fielder's choice, which of course doesn't show up in the box score as reaching base.

   Is there a way to re-generate this list to exclude events like that?

2. SpastikMooss Says:
   December 29th, 2009 at 2:06 pm

   Unrelated to this post, but I just wanted to say that I started using the index tool finally and I have to say that it's awesome. For instance, I just found out that Reed Johnson was hit by three pitches in a game on two separate occasions, almost exactly a year apart. You just can't make that stuff up.

3. Steve Lombardi Says:
   December 29th, 2009 at 6:31 pm

   SpastikMooss - agreed. PI is awesome.

   zuty - could be a glitch. I set the filer for Times On Base = Zero when I made the list. Then again, perhaps a FC doesn't count as being on base...it doesn't in the box score.

4. Raphy Says:
   December 29th, 2009 at 6:44 pm

   Not a glitch. FC don't count in the TOB. Neither does reaching base on a K+WP. This allows you to create some impossible sounding team batting game searches. Things like TOB=R with 1 runner left on or GDP=TOB with a run scored.

5. zuty Says:
   December 29th, 2009 at 7:15 pm

   Why don't these things count as "time on base"? (I've never known "time on base" as a legitimate stat, but I'm sure it is, or might as well be.) But why, if a player is actually on base, does it not get acknowledged? In truth, 50 players in a row could reach via K+WP, resulting in a bizarre first inning score of 47-0 with no one out, no one on base ever (but the bases loaded)...

6. DoubleDiamond Says:
   December 29th, 2009 at 11:01 pm

   And wouldn't being safe on an error also not show up in these situations?

   I remember a big deal being made over the first time Hideo Nomo became a baserunner - reaching on a failed sacrifice bunt attempt, in which the lead runner got thrown out. I guess that would also be considered to be a fielder's choice.

7. damthesehigheels Says:
   December 30th, 2009 at 2:38 am

   Zuty: there would be two outs in your scenario...

   But on that same note.. if Rose didn't get a time on base, shouldn't by that (flawed) logic, the runner he 'replace' get credit for the run scored?

8. Raphy Says:
   December 30th, 2009 at 4:43 am

   I believe that Times on Base is a made up stat by Sean for convenient use in PI. My guess is that his intention was to simply add all the recorded times a player reaches base (hit/bb/hbp/roe). The 2 plays that I listed before are never tallied (I don't think) and are therefore not included in the TOB stat. I wouldn't read too much into it.

9. Steve Lombardi Says:
   December 30th, 2009 at 9:29 am

   Raphy - thanks for the extra info on this one! 😉

10. Andy Says:
    December 30th, 2009 at 12:51 pm

    I have been involved in a lot of arguments about the Times on Base stat. My position is that reaching by FC should not count since most of the time the batter would have been out had the defense chosen to throw to first base, but that ROE should count since sometimes this happens due to speed/effort by the batter "forcing" an error. My logic is admittedly VERY lose and it's arbitrary...hence the arguments.

11. statboy Says:
    December 30th, 2009 at 1:37 pm

    Here's the Wikipedia definition.
    "In baseball statistics, the term times on base, also abbreviated as TOB, is the cumulative total number of times a batter has been awarded for reaching base as a result of hits, walks and hit by pitches. This stat does not include times reaching first by way of error, dropped 3rd strike, fielders obstruction or a fielder's choice, making this statistic somewhat of a misnomer."

12. tomepp Says:
    December 30th, 2009 at 2:21 pm

    I think that the TOB definition is just fine as is. Perhaps a better term for it might be "Times Earned Base". If we're going to credit a batter with a TOB for FC, ROE, K+WP, and other situations where he did not "earn" the base, why not credit him for a TOB when the manager inserts him in as a PR? The key to the stat is that the batter did something to earn passage to first base (and possibly beyond) and was not merely the beneficiary of miscues or decisions by the opposing team.

    Andy: you have an interesting point about forced errors, but in my anecdotal experience, those situations are only a small percentage of ROEs. In any event, it would be a judgment call as to whether the error was forced due to rushing or whether the error would have been made anyway. I have seen many incredible off-balance throws to nail speedy runners, and I have seen bad rushed throws made even when the runner was not a speed demon. (One botched throw to first when Hector Villanueva tripped on his way to first base comes immediately to mind...) I'm in Sean's camp on this - ROEs should not count. (Those speedsters who actually do force throwing errors simply get an extra chance to pad their SB and RS stats.)

13. JohnnyTwisto Says:
    December 30th, 2009 at 2:30 pm

    Reaching base on an error is absolutely a skill. Yes, it is due to a mistake made by the defense, but one could say that about a lot of base hits as well. There are players who consistently reach base on an error a lot (Derek Jeter has led the league five times), and players who don't. The player has reached base, no out has been made, whether it was called a hit or error is often just accounting -- the player deserves credit for getting there. Whether one wants to include it in a "Times on Base" stat depends on what one is trying to measure, I suppose.

14. tomepp Says:
    December 30th, 2009 at 5:40 pm

    Interesting point, JohnnyTwisto, but I would ask this: is Jeter's higher ROE total due to "skill" or simply more times making contact? It would be interesting to see year-by-year ROE/GBO (Ground Ball Outs) ratios and see if he - or anyone else - is consistently statistically significantly above the league average. One of the reasons that strikeouts are slightly worse than other types of outs (-.310 vs. -.299 in Linear Weights according to an article recently linked from a reply to a post in this blog) is that there is less chance of a defensive error. So making contact - at least in Linear Weights measurement - already credits the batter for the average chance of ROE.

15. DoubleDiamond Says:
    December 30th, 2009 at 6:19 pm

    I have noticed that every time I've seen a box score with the notation that someone reached due to catcher's interference, that catcher has also been listed as having made an error. I have decided that this has happened too many times to be a coincidence and that the way to "account" for someone reaching base on catcher's interference is to charge the catcher with an error. So, if tallying reaching on interference is now separate from tallying reaching on an error, in this case, the play is apparently counted in both categories.

16. JohnnyTwisto Says:
    December 30th, 2009 at 6:33 pm

    Tomepp, I'd assume in Jeter's case (and probably most players who consistently reach on a lot of errors), it's a combination of hitting a lot of grounders, being right-handed, and being fast. Does he reach on more errors than would be expected based on the number of grounders he hits? I assume so, but I have not checked. May try to look at that later. As for linear weights already crediting a batter for the possibility of an error, I need to think about that a bit.

17. JohnnyTwisto Says:
    December 31st, 2009 at 12:48 am

    I took a look at errors per groundballs + bunts in MLB in '09, '04, and '99, using the batting splits pages. Using those 3 seasons as a sample of Jeter's career, players ROE on a grounder or bunt about 2.7% of the time. Jeter has ROE on a grounder or bunt 3.7% of the time. May not seem like a big difference but it is almost 40% higher than average.

    As for the difference in the value of a K and a non-K out, I'd think most of that is due to the runner-advancement of non-K outs than of ROE. You do make a good point that linear weights which count ROE as outs should probably not have another value for ROE. But of course the raw numbers we look at on here, like Times on Base, are not based on linear weights at all.

18. tomepp Says:
    December 31st, 2009 at 10:59 am

    Interesting results, JohnnyTwisto. Yes, I would say that 3.7% vs. 2.7% is probably statistically significant (though you'd need a larger sample size to accurately determine statistical significance). Did you select the years you did for Jeter at random, or was there a reason for selecting those years? I could not find GB and FB data for batters on their B-R pages (other than GB/FB ratios, but that would only give an approximation when multiplying through for career totals), but I did find GO and AO totals for the years 1999 - 2009 on the player's mlb.com stats pages. Where did you find the GB totals for Jeter? To find his ROE numbers, I had to go to each year's splits pages separately. (Note to Sean, et. al.: could ROE totals be added in a column after the IBB column on the player's base page - or at least on their "more stats" page - rather than being exclusively on the splits pages? It really has nothing to do with splits...)

    The results I got for Jeter for ROE/(GO+SH) from 1999 - 2009 is 112/(2251+61) = 4.84%, which is higher than the values you got, but I'm guessing you counted GB hits in the denominator. Unfortunately, that is as far as I could go, as I could not find league ROE totals anywhere.

19. JohnnyTwisto Says:
    December 31st, 2009 at 1:34 pm

    Sorry I was unclear. I just looked at 1999, 2004, and 2009 when estimating the ROE % for all major leaguers. No reason behind it except that the span covers most of Jeter's career and I didn't want to calculate it for every season he's been in MLB. However, for Jeter himself I used his entire career, NOT just those three seasons. For both the league and Jeter's numbers, I got the data from the "Hit Trajectory" section near the bottom of the splits pages (http://www.baseball-reference.com/players/split.cgi?n1=jeterde01&year=Career&t=b - I don't see a direct link to that section). That gives numbers for groundballs, flies, liners, and bunts. I just PA for the total number of grounders and bunts, and ROE are listed there as well. So yes, hits are included in the denominators. Maybe they should be removed? Let me try it that way, just ROE as a % of groundouts/bunt outs:

    Jeter is at 5.0% for his career. MLB for those three seasons is at about 3.5%.