Did the right pitcher get the win?

This just happened to my son (relief pitcher) and I questioned that he got the win.  In a 9 inning game, the starting pitcher has to pitch at least 5 innings to get the win. It is discretionary to the scorekeeper as to who gets the win based upon which relief pitcher was most effective. Since the RP#1 pitched two innings, regardless of how effective he was not, I bet the scorekeeper gave him the win because he pitched two innings and the team was still up by 2 at the end of the game.  I think I got this right, but I always stand ready to be corrected. 

I don't know. Pitcher Wins are stupid. It's not a real stat. Don't pay any attention to it. Baseball statistics are things that actually occur on the baseball field. Pitcher W/L does not occur on the field, it's a made up construct with arbitrary boundaries that has no reflection on an individual pitcher's performance. Don't worry about it.

I agree with JH in that the win/loss stat does not accurately reflect the sole performance of the pitcher.  But it must mean something to some folks because it is always listed by a pitcher's name when they are mentioned in an article, etc.

You're correct, but that meaning is entirely misplaced. The practice of citing W/L should, and will, eventually change.

Somehow thread drift set in almost immediately.

The question wasn't the philosophical debate abput the efficacy of wins or losses. The question was, under the present rules, did the right guy get the W. Fwiw, I agree with JHs philosophy.

I agree with Southern mom and believe that she gave you a correct and detailed answer.  After that, what the heck, why not drift?

My answer was I don't know, and my reason was because Pitcher W/L are not a real stat. That was me answering your initial question, not drifting the thread. I honestly and truthfully have no idea, because I don't look at Pitcher W/L. It isn't real and shouldn't matter.

I am glad you agree though…a few years ago, my post may have started an uproar. 

JH,

I don't believe that I will see it in my lifetime that W/L are not mentioned!!!

RFF- Keep your young self healthy and live a long life…ya never know.

Here’s where things get a bit goofy. If you read OBR 10.17 closely, the OSK is given some latitude to use his judgment. Without knowing a lot more about that particular game, it’s really not possible to get a feel for that judgment.

FI, it may be that RP #2 faced the top half of the Detroit Tigers lineup or the bottom half of the Met’s lineup. If it was the former, he did very well. If it was the latter, he didn’t do well at all.

In the end, though, like most of the others, I agree that pitcher W’s & L’s are really a pretty lousy way to show much of anything. Unfortunately, like BA, it’s something ingrained into the narrative of the sport, and as long as it’s something required and defined by the rules, its gonna stick around.

If you want to kill Wins/Losses, get the folks at GameChanger to drop it. When you finish scoring a game, that's the first thing it wants to know.  With GC going strong in LL, travel, and HS that gives a whole new generation of scorekeepers the idea that it's very important.

Originally Posted by JCG:

If you want to kill Wins/Losses, get the folks at GameChanger to drop it. When you finish scoring a game, that's the first thing it wants to know.  With GC going strong in LL, travel, and HS that gives a whole new generation of scorekeepers the idea that it's very important.

I assure you that if MLB dropped assigning winning and losing pitchers and making it a part of the official scorer’s report to the Office of the Commissioner, Game Changer would drop it as soon as they could get a programmer to do it.

Its far more likely that over time the numbers geeks will use it less and less, the same way they’re using BA and ERA less and less, and it will go the way of all other dinosaurs. It may morph into something different and more meaningful, or it may just go away because it really is meaningless.

What numbers baseball uses is really a pretty dynamic thing. FI, if you go to baseball-reference and look up Quality Start, and in the narrative pretty much says it’s a better stat than wins or losses because it’s a lot more objective. But then it goes on to say that Game Score is a metric finding more favor with Sabermetricians. Heck, back in the 70’s no one heard of WHIP, now it may be just as popular as ERA.

The thing is, stats that actually mean something and help make decisions will always be the ones the real baseball people use, while the stats the most people see will always be the most popular.

Referring to the original question, it is a judgment call, and all four relief pitchers deserve some consideration:  RP#1 got out of the jam in the 5th inning when the tying run was on base,  RP#2 got the most outs of any reliever, RP#3 was flawless, and RP#4 nailed down the win.  

Meanwhile, the starter--who did the most work, retired more batters than everyone else combined, and gave up no runs--doesn't even get considered.  This game is a great example of how meaningless and arbitrary W-L is.  Doesn't matter who you give it to, the most deserving pitcher won't get it.

The sooner we rid ourselves of W-L the better.  There are so many times the "wrong" pitcher gets a win (e.g., the closer who blows a save in the top of the ninth and is rewarded with a win because the offense scores two in the bottom of the inning) or a loss (e.g., the pitcher who gives up an unearned run and gets no run support).  We should replace it with some agreed upon "slash line" that gives a measure of pitching performance that relates to not letting batters put the ball in play, not letting them get on base, and not letting them score--things that happen on the field that contribute to minimizing runs allowed.  

I would also de-emphasize ERA because a) it encourages the belief that it's somehow ok to let a run score after an error when in fact after an error is exactly when a team most needs its pitcher to be stout and prevent the run from scoring, b) it often rewards pitchers who don't do their jobs, such as the short reliever who comes in to a bases loaded, two out situation, gives up a 3-run double, retires the next guy, and sits down with a perfect ERA.

Swampboy- Good post. I'd encourage those interested to read up a little on FIP. That's probably the best measure of pitcher performance available in the public spectrum. It's not perfect, but it's much more accurate than ERA, WHIP, or anything else that involves factors not entirely in the pitcher's control.

There are batted ball variables that are within a pitcher's control to a certain extent, but without proprietary measures available to MLB teams, the public is not able to gain access to such results. For now, I'd highly recommend to coaches and players that are trying to learn how to best analyze pitching to follow FIP.

I took JH up on his suggestion and found this to be the most thorough quick read on it...

http://www.fangraphs.com/library/pitching/fip/

Just looked into FIP.  Looks like this is considering HRs, BBs, HBPs and Ks as the only things that are meaningful to a pitcher's effectiveness.  These numbers are plugged into a formula:

FIP = ((13*HR)+(3*(BB+HBP))-(2*K))/IP + constant

in which the constant is the same formula (without the constant), using numbers for the whole league:

FIP Constant = lgERA - (((13*lgHR)+(3*(lgBB+lgHBP))-(2*lgK))/lgIP)

So, what you are really doing is looking at things that are ONLY in the pitchers control and then comparing what you do against the league average.  You're doing good if your FIP is below the constant (league average) and not good if your FIP is above the constant (league average).

Here's what they say about the weights given to the different categories - "The individual weights for home runs, walks/HBP, and strikeouts are based on the relative values of those actions with respect to run prevention." and I'm not sure where they came up with that.

This is OK, but it seems to me that there are other things that help determine a pitchers effectiveness other than HRs, HBPs, BBs and Ks.  There is always talk about how ground ball outs are much better than fly ball outs for example.  There will always be balls put into play and to not consider the mix or even consider that balls are put into play for a pitcher in measuring his effectiveness, leaves a lot out of the equation.  

This may be one tool to use in measuring the effectiveness of a pitcher, but I don't think it can be the only one.

You could potentially have a pitcher with a VERY low FIP if he doesn't give up many HRs, has low BBs and a lot of Ks.  But what if he gives up a TON of gappers and gives up a lot of runs.  Not sure I'd want that guy.  We had a guy that would have outings like that on one of my son's teams.  He would strike out the side, but would give up 3 singles and two doubles in between the strike outs.  FIP would not take any of this into consideration.

bballman- The weights in the FIP formula were derived from simply taking the run value of each action for each season (http://www.fangraphs.com/guts.aspx?type=cn). As you may infer, the run scoring environment at levels aside from MLB could be very different, and therefore the weights may change (as would the constant, given the league-wide necessity of the actions). Using the provided formula would give you a sense of production, but wouldn't be entirely accurate at other levels.

To address your point on batted balls: there is evidence that pitchers have some control over how well hit a ball is, as well as if a ball is a ground ball or a fly ball. While this can be determined using proprietary technology, the results of this technology are not publicly available. However, the example you provided does not eliminate the importance or accuracy of FIP at all. That same pitcher you mentioned could have gotten 5 outs on hard hit balls instead of them going for hits. Pitchers have absolutely no control over exactly where a ball lands in the field. I assume the pitcher you're referring to probably threw a maximum of about 100 innings or so (given a typical collegiate workload). My former coworker, Russell Carleton, did a study analyzing at which point statistics stabilize and no longer become a small sample size (AKA: irrelevant). In the case of batted balls, the statistic in question is "BABIP" - Batting Average On Balls In Play (I.e. How many batted balls in play fall for hits). Russell found that the stat typically stabilizes at around 2,000 Balls In Play (http://www.baseballprospectus.....php?articleid=20516). So, while your son's teammate may have been unlucky in giving up hits in places fielders weren't, the batted ball results against him would most likely regress over a large enough sample to the point where his FIP is much more closely indicative of his True Talent in comparison to his competition. In layman's terms: Luck sort of evens itself out over time.

FIP certainly isn't perfect, and it certainly isn't the only measure. There's no such thing as a perfect statistic right now. But in the public sector, it's the best one available to measure pitcher performance.

Let me know if you're interested in hitter performance - wRC+ is my favorite baseball stat.

Will do JH.  And this statement is very true as well.  While one pitcher may give up some hard hit balls that happen to find gaps or holes and time would allow those hard hit balls to go right at the defense later down the road, the same thing could be said of the balls that are hit weakly and happen to find a hole.  The seeing eye single, the bloop just over the infield, etc.  

As a pitcher you know it goes both ways.  Sometimes you'll have a bunch of balls hit hard off you right to someone and the weakest hit balls go for hits.

I agree, this isn't a bad measurement, however, I just think there are some holes.  The problem is, I just don't know if there is a totally accurate, purely statistical way to quantify a pitcher's effectiveness.  Baseball is a team game and no matter how hard we try to isolate how good a pitcher is, at some point, every pitcher will have to rely on his defense.  I guess it doesn't hurt to keep trying to find that magic formula, but I fear when it is found, it may be so confusing no one will be able to grasp it or use it.   

bballman- For MLB teams it's important to attempt to best quantify a player's True Talent because there's a lot of money hinging on that player's future projected performance. No doubt that defense is extremely important, and luck is involved, but as organizations that must pay players for their services, it behooves a team to have a grasp of the ability to eliminate variables outside of that player's control. For example: If a hitter has a high BABIP for one season, a team may investigate why that is the case. Is it because the player has above average speed and beats out balls other hitters can't? Is it because the player hits a lot of line drives and makes consistent hard contact? How much of these variables factor into luck? Will the player regress to the point where his offensive performance will decrease? 

Those are just a few questions that need to be asked. In order to get to the closest possible answer, teams need to use advanced metrics to find associating measures of True Talent. It's still kind of a crapshoot but without those metrics, it would be even more of a crapshoot.

As for a "perfect" formula - I don't think there's such a thing. We're still dealing with human beings, with emotions and personalities and ailments. But I guarantee that some of these stats I'm citing will be ancient a decade from now. Formulas are created by humans, so there's no such thing as a formula people can't understand.

Oh, yeah!  I'm liking this stat.  Played with my son's numbers from last spring.  I'm having to do some fudging on league data since his conference stats leave out some key elements, but no matter how I slice the league info, his 1 HR and 9 BB's in 62 innings will make him look a lot better than his 3.34 ERA does.

Josh, please tell me scouts care about this stat!

Haha. Well as I said the constants change per level…league average HR%, for example, is something I would imagine is lower in the NCAA than in MLB (I don't know for sure, just guessing). 

However you slice it, striking out a lot of players and not walking a lot of players is a good thing. A formula isn't needed to tell anyone that.

I don't know about other scouts, but I'm on Fangraphs everyday and I used to write for Baseball Prospectus. So, I care...

I’ve looked into producing FIP for our pitchers several times over the last several years, but I run into the same problem with it that I run into with so many of the other metrics used for MLB players. The constants won’t work because there’s no way to get them. L Remember, this is HSBB.

Even though the article tells how to compute the constant, and it is pretty simple, when these guys talk about the league ERA, they’re talking about AL or NL. In our section of Ca, we have 195 schools and 27 leagues, playing in 7 different divisions. We play about half our games outside of our league, at least 3 or 4 outside of our section, and at least 3 or 4 outside of our division. Then to make things even more difficult, not all teams post all stats from all games, so that constant is really a PITA to come up with.

So here’s what our team’s pitchers look like. The 1^st page is all the pitchers and the 2^nd is the pitchers who’ve thrown at least 25% of the highest number of innings pitched since 2007. I threw 3 in as a constant for no particular reason, but for me, not having the constant is just fine if I’m comparing only our pitchers and ordering the results by the FIP # without the constant.

I do think numbers like this one and others could certainly be useful to scouts looking at HS or college players. Not that they’d make it possible to pinpoint a pitcher who would make it at the next level, but they could certainly be used as another “filter” to help them get through the hundreds of thousands of players in the mix.

Back to the original question, based on the information provided, I would give the W to RP #1.

All these stats are great, but the scorer needs to decide before he closes up the book at the end of the game. I'd probably nibble on my pencil for a minute or two before deciding.

The worth of W/L for relief pitchers is dubious at best and agree with all the discussion. I don't think we'll be seeing this statistic go away, because it is somewhat meaningful over a starter's career (starters with 300 wins and HOF). For relievers, I've seen guys with a few wins that were achieved by blowing the save and then having his team come back after he's gone.