Thursday, May 26, 2011

Jo-Jo's Circus

NEW YORK -- Every now and again, you will hear somebody make fun of baseball numbers geeks by inventing some silly made-up statistic of their own. For instance, they might say: "Hey, did you know that Player X is hitting .403 with runners in scoring position on Wednesdays when the starting pitcher is left-handed, a Sagittarius and has broken up with his girlfriend within 1,093 hours of taking the mound."

LOL! See, the joke is that baseball stat-heads measure obscure and pointless things. The Simpsons pulled off this little joke in their recent baseball show, but the truth is that the joke has become ubiquitous. You can hear it on radio, on television, in newspapers, on Broadway, in ballparks across America. Everyone who wants to poke fun at baseball stats people come up some version of the joke.

"I'm not interested in the guys batting average against lefties on national holidays when the person eliminated on American Idol has a name that starts with B."

"Oh, what is he hitting in domes during day games when the NASDAQ is down?"

"Yeah, they're in their mother's basement figuring the pitchers wDIDDY on turf when he batter scored better than a 1000 on his SAT."

And so on. The jokes can be blandly funny, I suppose, though I think they miss the point. There ARE some annoying numbers geeks out there who chase after utterly meaningless statistics. Those geeks, though, might not be the ones we generally point at.



What I mean is that the people who spend a lot of time breaking down baseball numbers, well, their methods may be baffling to those of us without much feel for math but in general they are working to find things that could be really interesting. There are people out there who work hard to come up with mathematical formulas to determine the run values of different actions (how much more a single is worth than a walk, for instance). There are people out there who work the numbers to separate the pitchers contribution in run prevention from the defense's contribution. Some try to break down the statistics to see if certain players have the unique talent to hit better in the clutch than they do in regular situations. Some try to pick through the records to see if pitchers -- individually and collectively -- can pitch to the score. Some just work through the numbers to find counterintuitive facts -- for instance, the team that leaves more runners on base tends to win more often than lose. You might not think that.

Sometimes, these mathematical efforts go over my head. And sometimes they go WAY over my head. But the point is that much of baseball number crunching is for a purpose -- to answer a question, to prove a point, to discover a whole new way to look at baseball -- and that can be fascinating if you come at it with an open mind.

Then there are really and truly pointless baseball numbers, statistics that tell you almost nothing, records that are made to be insulted. What would be a good example of a bad baseball statistic?

Well ... if you type in Toronto pitcher "Jo-Jo Reyes" and "winless streak" into your Google News Feed, you will find one. You will actually find hundreds and hundreds of hits representing one. Yes, this "streak" -- and you will notice I put the finger quotations around the word --has earned its corner of the news. People in the mainstream are actually talking about Jo-Jo Reyes' winless streak. And to be blunt about it this record is dumber than most of the fictional jokes people invent just to make fun of baseball.

To be clear: After Wednesday's start against the Yankees, Jo-Jo Reyes has now made 28 consecutive starts without earning a win, which apparently ties him with Matt Keough for some sort of dubious distinction. But what distinction? That's a bit trickier. Is his record 0-28? No, it is is not. It's actually 0-13. Is his team's record 0-28 in his starts? No, it's not -- it's a much-less interesting 7-21. Does this streak go back to last year? Well, no, because Jo-Jo Reyes didn't make a big league start last year. He only made five big league starts back in 2009. So this "streak," and there are those finger quotes again, goes back to 2008, when he was a kid playing for Atlanta. He did not win any of his last 13 big league starts that year. He did not win any of his five big league starts in 2009. And, of course, he has not won any of his big league 10 starts this season.

That's not a streak. That's bad timing. That's a not especially intriguing coincidence. A streak is something that happens continuously. A pitcher who goes up and down, makes a spot start here and there, goes to the minors for most of one year and the whole of the next, no, that's not a streak.

But the quirky timing of the streak is not even the biggest problem with this bit of nonsense. The biggest problem is that it's all made up. This whole thing is built around the archaic win statistic. If the Blue Jays had lost 28 straight games Reyes had started, yes, that would be interesting. If Reyes had left 28 straight games with his team trailing, I could see THAT being interesting. But neither thing is even remotely true. Heck, in Reyes' LAST START he threw seven shutout innings and the Blue Jays pen blew the game. The start before that, the Jays actually won the game but because of the goofiness of the statistic, Reyes did not get the credit. Those are his last two starts, for crying out loud.

So, what, he hasn't gotten one of those outdated "pitcher wins" things? Big deal. I've got a long, long streak going where I have not used the word "mollycoddle*." I don't think presses will be stopped.

*Damn. Streak broken.

I rip on the win as a statistic pretty often here, but if I'm being honest, I would miss the win if it was suddenly gone. I think it adds to the appeal of baseball to be able to say a pitcher is 17-2 or 4-15 or whatever. Wins are an easy shortcut and a common language between casual and intense baseball fans. And in concert with other statistics, wins can give you a decent feel for what kind of year the pitcher is having -- whether it's a good year, a bad year, a lucky year, a hard-luck year, whatever.

But this "winless streak" annoys me because it attempts to give an authority to the statistic that is so obviously not there.

After Wednesday's non-win -- Reyes made sure on this one by giving up back-to-back doubles to start the game and five runs in three innings -- reporters surrounded the man of the hour. The whole scene was bizarre. Hey, here's one of those kooky joke stats: What pitcher has made the most consecutive starts without pitching at least five innings and exiting with his team in the lead in a game that the team ends up winning?

And people make fun of BABIP. Sheesh.

19 comments:

  1. I think Reyes as a #5 starter has done a decent job this year. The majority of his starts he has kept the Jays in the game. And he also has the second highest BABIP in MLB this year and isn't walking a pile of guys.

    ReplyDelete
  2. Patient: My back hurts in four different places.
    Doc: Well, stay out of those places.

    That is the level of humor achieved by the made-up stats jokesters. Keep fighting the good fight, Joe.

    ReplyDelete
  3. The real blame goes to the stat guy who dug through all the numbers and came up with the stat that Jo Jo Reyes hasn't won in 28 straight starts.

    ReplyDelete
  4. It doesn't annoy me - it's an oddity, like the unassisted triple play, or hitting for the cycle.

    ReplyDelete
  5. I actually disagree with you here Joe. It seems to me that the attention to this streak has not been focused on how bad Reyes is, but rather on the confluence of factors that had to occur in order for this streak to be maintained over the 4 years (as discontinuous as that may be). In that regard, it has been suggested, it is quite remarkable (or bizarre), even if it does not tell us very much about about Jo-Jo Reyes as a pitcher. This is most likely true. Let's assume that about 35% of games started end in the starting pitcher getting a victory (I have not done a thorough analysis of this, but it actually seems like a pretty accurate estimate after skimming through how many wins pitchers tend to get given number of games started and total games pitched). Using this assumption and assuming Jo-Jo Reyes is an average pitcher, the chances of him making this many starts without getting a win would have a probability of .65^28 or 0.00058% or 1 in 173,151. However, perhaps my estimate is too high, and it is probably accepted that Jo-Jo Reyes has been a below average Major League pitcher in this stretch (as evidenced by long stints in the minors). Therefore, let's up the stakes and say he has only a 20% chance of winning any given game he starts. The chances of him making this many consecutive starts without a win would be .80^28 or 0.19% or 1 in 517. Now consider the fact that he has had the opportunity to make that many starts (even over the course of many years), and this is actually a very rare event. Like it or not, wins are still a part of baseball lexicon. As far as bizarre streaks involving wins go, this streak is pretty unusual...which I think makes it a story.

    ReplyDelete
  6. In a certain technical way, Joe, your streak is still alive. You didn't use the word "mollycoddle," you mentioned it.

    http://en.wikipedia.org/wiki/Use%E2%80%93mention_distinction

    See?

    ReplyDelete
  7. Listening to Royals games this year, both teams of Ryan and Frank and Denny and Bob seem to continually comment on the theme of bullpens having greater and greater impact on the game. It's funny because while that's irrevocably true I don't think it's any more true this year than it was 3 years ago or 8 years ago. Over say the last 20 years, though, bullpens certainly have come into greater use and the use of bullpens has made the pitcher win particularly useless.

    I'd be interested to see what the decision percentage is among MLB starting pitchers.

    ReplyDelete
  8. It's even more convoluted than that: Which pitcher has made the most consecutive starts without pitching at least five innings and exiting with his team in the lead in a game that the team ends up winning without ever relinquishing the lead?

    ReplyDelete
  9. I was not only at Jo-Jo's latest non-Win, I spent a couple innings arguing sabermetrics with a couple of Yankee fans who think stats are just a bunch of mumbo-jumbo Mets fans cook up to take down their superiors (the topic was, natch, Jeter). But there was a moment of light when one of them asked about that OPS thing up on the scoreboard, and allowed me to explain why Mark Texeira's .253 average is so much better than Jeter's .257.

    ReplyDelete
  10. I hate Wins. They mean nothing. They're an artifact of ignorance, and when a pitcher's win-stats come tripping off my tongue I have to stop, mid-sentence, and correct myself.

    Let the wins go, Joe. You know -- for the kids.

    ReplyDelete
  11. Denopac said: "It's even more convoluted than that: Which pitcher has made the most consecutive starts without pitching at least five innings and exiting with his team in the lead in a game that the team ends up winning without ever relinquishing the lead?"

    And then you have to add "... or at least four innings in a game that lasts five innings on defense..." as well as "...not including official minor league games that were started and won in the middle of the streak."

    ReplyDelete
  12. @keatang, you mean those particular Yankees fans, correct? I have spent many nights at the stadium trying to explain exactly that.

    ReplyDelete
  13. I seem to remember that you did the research and discovered the Simpsons joke was actually completely wrong... One of the few times the Simpsons writers have failed us!

    ReplyDelete
  14. The stat's even more convoluted than you make it: What pitcher has made the most consecutive starts without pitching at least five innings and exiting with his team in the lead in a game that the team ends up winning--and without the pitcher's team having given up the lead after he has left the game?

    ReplyDelete
  15. To me, the summer I was Sports Editor of the Indiana Daily Student, the biggest news story wasn't the Barcelona Olympics, but the Anthony Young losing streak.

    ReplyDelete
  16. This reminds me of something I was thinking about today --

    Question: What percentage of sports writers/journalists are:

    a) Overweight
    b) Bald
    c) Have a goatee
    d) All of the above


    I watched ESPN today for the first time in years -- why hello there, long holiday weekend! I actually ended up watching for about 40 minutes because I was waiting to see one face on the screen that did not meet one (but, for the most part, multiple) of the above criteria. I quit after about 40 minutes because I became bored with the game. But, seriously. Man, oh man. I've heard of Groupthink. But this is something even more advanced than that.

    ReplyDelete
  17. The non-stathead casual fans' criterion for stats is simple: If the media they watched/read when they were kids used the stat, it is meaningful. If they didn't, it is a joke stat. By that standard, for most people, wins are meaningful and BABIP gets tossed in the joke bin with American Idol and NASDAQ.

    But despite all the jokes, the attitude to stats nowadays is orders of magnitude better than it was, say, 20 years ago.

    For all of you young statheads, you have no idea how much better you have it now than we did 20 years ago. If it were 1991, *no one* you knew would pay attention to OBP and SLG instead of BA, HR, RBI. No one you knew would adjust for park effects. The progress in the general knowledge of SABR stats in the last 20 years has been astounding. It will continue. The jokes will get old.

    ReplyDelete
  18. Great, great, great, great, great piece. Just great.

    ReplyDelete
  19. Last night, I read this while watching the Pirates-Phillies game. Jeff Karstens pitched 7 great innings. Cole Hamels pitched 8 great innings.

    Reliever Daniel Moskos threw 8 pitches in the 12th inning.

    He got the win. Great column, Joe. As always.

    ReplyDelete