Somethin about this stat seems to mystify me.

I was lookin at Lee Smith's CAREER stats today, and over his career, he had (only) a WAR of 29.

Really?

He was only good for 29 wins over the average relief pitcher, in his entire fuckin career (18 years)?

Seemed hard to believe. I mean thats like saying his average replacement would have had about 2 saves per year less than Smitty. Which in itself woulda still been great. Also, I would think stoppers have a higher WAR than some starters., since if they save 40 or 45 games a year, they are involved in more wins.

By contrast, Rizzo ALREADY has a WAR of 31.5 for his career. Sounds about right. A value of 4 or 5 wins per year for one player is about right for a good player like him, on a good team full of good players who have good WARS in their own right.

So somebody explain this WAR to me. Does 29 sound about right to you for Smith?

Do u think WAR is a useful stat? I mean it seems like a good concept to me, but not sure if its being applied correctly.

_________________

Mariano Rivera had a career WAR of ONLY 56.

Really?

652 fuckin saves, a winning record (82-60), and a 2.21 career ERA and you only have a WAR of 56?

I would think for him that should be around 150.

I mean, fuck, if u turn some of his saves into blown saves (leaving him with still, for example, 600 saves), shouldnt that theoretically kill his WAR pretty quickly?

_________________

What is it good for?

Absolutely not evaluating the value of relievers

_________________
"Now comes good sailing. Moose. Indian."

The thing about saves is that rationally, 2-3 run leads should almost never be blown.

I found a site online that calculates average runs expected in different situations (https://gregstoll.dyndns.org/~gregstoll ... nning.html), and taking what it spits out for coming in at the beginning of the 9th inning with an average pitcher and an average hitting team:

- There should be a 87% chance of saving a 2-run lead
- There should be a 95% chance of saving a 3-run lead

So closers really have a hard time adding to their stats in those situations.

Of course, saves are not rational, because of nerves. Look at Carl Edwards or LaTroy Hawkins. But a rational model like WAR or WPA+ doesn't understand how to adjust for that.

For closers, I think there are only two good metrics:
- 1-run saves converted/blown
- 4+ out saves converted

Giving up a run when you come in with a 3-run lead against the bottom of the order and you're experimenting and one goes out of the park, that's just as irrelevant as Sammy Sosa's multitude of 8th inning solo home runs when they were down by 4-5.

_________________
"When people want their version of the truth, they go find it, no matter how baseless their beliefs." -- Ken Rosenthal

ERAs are often deceivingly good for relievers because they sometimes come in with one or two outs in an inning.

Back in the day, Murph suggested having a modified ERA system where runs allowed would count against both the pitcher who let the runner on and the one who let him score. Let's say the first pitcher allowed a runner to reach first, and then the reliever allowed him to score. The first pitcher would be charged with 1/4 run and the other pitcher 3/4. That made a lot of sense to me...I assume some of the latest metrics do something similar to equalize a reliever's performance and also reflect the fact it's easier to keep your ERA low when you only have to get one or two outs.

_________________

I would argue it excellent for that very thing. "Closer" is the most overvalued position in the game.

_________________
General Jim wears Fort Bragg down... to get it!

I would agree with the second part. The issue is that WAR is a compiling statistic. Relievers only pitch ~70 innings per season, so an elite reliever can often fail to accumulate more WAR than a mediocre starter.

_________________
"Now comes good sailing. Moose. Indian."

But that reflects- correctly- the limited impact of the reliever.

_________________
General Jim wears Fort Bragg down... to get it!

Until you see a bad one! A bad closer can just kill a season.

(Not sure if you mean closer or relievers in general.)

Is that perception or reality though? Of course when you lose a game in the ninth inning it stands out more than when you blow it in the fifth.

I agree that the last three outs seem to be difficult. And it appears that certain guys do not have the makeup to consistently and successfully get those three outs. But whether my perception is based upon small samples and anecdotes, I can't be certain.

_________________
General Jim wears Fort Bragg down... to get it!

Honestly not sure. I was thinking about going back to look at a few Sox teams where bad closers cost them games, and I was wondering to what extent this was actually true.

On the other hand, I think a good closer can make a decent team really good.

Not all innings pitched are equally valuable. Mike Leake getting three outs with a 5 run lead isn't as important as Josh Hader coming in to strike out the side with the bases loaded. Yet, both go down as scoreless innings.

_________________
"Now comes good sailing. Moose. Indian."

I agree. But there are often high leverage innings pitched by relievers prior to the ninth. Coming in to a game in the fifth with 2 on and 1 out and coming away scoreless is probably worth more than coming into a fresh ninth and finishing it.

_________________
General Jim wears Fort Bragg down... to get it!

Adjusted WPA/out would do it, but that would lose too many people mathematically. But even then, you come in and pitch a clean 9th in three run game, you've added a whole 0.03 WPA over 3 outs, or 0.01/out.

_________________
"When people want their version of the truth, they go find it, no matter how baseless their beliefs." -- Ken Rosenthal

_________________
Sex isn't dirty, sex isn't a crime. It's a loving act between two or more consenting adults.

-Hank Kingsley

STOPPER does not equal CLOSER

_________________

If we're talking about changes to the game, one thing that I believe is a positive development over the past couple seasons is more first pitch swinging. The first pitch is often the best pitch of an at-bat and if you're taking it in the interest of merely "seeing pitches" you're helping create a boring game dominated by walks and strikeouts. It got to a point where pitchers could impudently throw a get-me-over fastball on the first pitch and start many counts 0-1. The increase in first pitch swings is clearly an adjustment to this.

Another thing I think is interesting is that all these teams have their detailed spray charts and proprietary data that informs their shifts, yet BABIP remains right at .300 no matter what the defense does. Tip of the cap to Voros McCracken.

_________________
General Jim wears Fort Bragg down... to get it!

fuck!

_________________

whats BABIP mean?

I'm still stuck at BA, OBP, and SLG, only.

_________________

I just find it hard to believe that Rivera, a legend, only had a value of 56 victories over the average replacement.

I mean he came in a lot leading by only a run, came in with men on base in the 8th inning, etc. Did more 4 and 6 out saves than most of his contemporaries, in an age of specialization.

As far as Lee Smith, he pitched a ton of innings for a reliever. Up to 140 IP some years with no starts. So he was in games longer and therefore had a LOT of decisions for a reliever, and he was well under .500 for his career. So, with so many losses I can sorta see where it dragged down his WAR to negate the high save total.

But Rivera, 82-60, 652 saves, 18 years, I woulda thought he would be up around 80 or 90 for WAR.

_________________

WAR has always been a flawed stat designed from the beginning to find the answers that were wanted. It starts off with a bunch of assumptions that fall apart so much people have to say "Well, it's only good for comparing players who play the same position" even though a positional adjustment is one of the whole points for this. It then throws in defense without adjusting for how little defense matters at some positions.

It is an easily digestible stat for internet arguments and sports radio because it puts a number value on a player so you can say "No, Player X is better because he has a WAR of 3.1 and Player Y has a WAR of 2.0!" without having to think about anything else in that equation.

_________________
You do not talk to me like that! I work too hard to deal with this stuff! I work too hard! I'm an important member of the CSFMB! I drive a Dodge Stratus!

the moment a "should have" factor is included into an equation, it no longer becomes scientific

_________________
O judgment! Thou art fled to brutish beasts,
And men have lost their reason.

You both are right. The part Rick makes about it being easily used in conversations and talk radio screams that point. The stats like that are fine if kept in the perspective necessary you can have both worlds and enjoy the game.

For me, it seems a lazy tool as well as sometimes flawed for people in fantasy or talk jobs that don't want to watch every player and/or game.

_________________
Elections have consequences.

_________________

Batting Average for Balls In Play.

Even though "batting" is in the title, it's really a pitching stat. The theory is that a pitcher can only control three outcomes: strikeouts, walks, and home runs. All other balls that are put in play are considered good luck (out) or bad luck (hit) and not a function of a pitcher's skill. This is because someone discovered that all pitchers at the MLB level seem to have about the same BABIP. Assuming that's true, then all you would need to do to judge a pitcher is look at his strikeouts, walks, and home runs against. You can also look at a pitcher's recent BABIP to see if he's been lucky (his BABIP is lower than average), or unlucky (BABIP higher than average) because over time it should come back to the average.

Of course, many disagree with the notion that a pitcher has no effect on whether a ball in play is a hit or out. Some pitchers would seem to be good enough to induce weak contact, while others get pounded.

BABIP doesn't work as well as a metric for batters, in that there is more variability depending on the hitting style. For example, Ichiro had a very high BABIP because of the many infield hits he had.

_________________

Yeah, it's a pitcher's stat. Theoretically, anyone who pitched a large enough sample would have a BABIP right around . 300.

_________________
General Jim wears Fort Bragg down... to get it!

I was never a Cub fan, but Lee Smith was my favorite Cubs player of all time.

When he came in, you knew the game was over for the other team. He threw fastballs past everyone.

Harry would get all worked up about him. It was very exciting to watch as a youngster.

I felt the same way. Exciting, and his slow walk was hilarious and I always imitated it.

_________________

So Harper turned a 2 run loss into a 2 run win with one swing o' the bat. Instantly. In the 9th. Walk-off.

In this case, does that increase his WAR by a full 1? Or at least close to 1, say .99?

I mean, very few players woulda hit an instant grandslam in that situation. Sure, some guys mighta hit a double, but that only ties that game and doesn't guarantee a fuckin victory.

So to the original question, does this hit increase his WAR by 1?

_________________