Build a Better WAR Metric, Part 2
Ok, you guys have spoken, and you don’t want a bases loaded walk to count the same as a solo HR. That even though the base-out state before the event and after the event remain unchanged, and that the number of runs now in the bank are the same, the WAY it happened matters to most of you. Therefore, we are NOT trying to preserve the runs, we are not trying to make sure the runs add up. You have been clear on that.
Now, let’s talk about “preservation of wins”. It’s a 0-0 game, the bottom of the 9th, the bases are loaded with two outs. Historically, at this point in the game, the batting team would end up winning 68% of the time. It’s a high stakes situation, a Leverage Index of 6.4. And the batter walks. The batting team wins, game is over. Ooops, I meant the batter hit a single. No, wait, it was a Grand Slam. No, wait it should have been a Grand Slam, but Robin Ventura decided to abandon the bases after he reached first base. Regardless, the game is over, and the batting team won as soon as the batter touched first base.
Your question:
given the literal meaning of the term ‘wins above replacement’ and it’s broadly context dependent perspective (playing time/position/leverage/etc), how could any of these cases be of differing value? the displacement in terms of win probability added of any of these acts is the same.
i don’t mean to say that there shouldn’t be an otherwise context neutral valuation of leveraged performance, but it seems like that should be a new metric. refinement should not redefine
Current WAR is largely context independent. Even openWAR, that uses run expectancy (which is context dependent), doesn’t take into account the leverage situation of the game.
I feel like the hitter shouldn’t be penalized due to the singular timing of the event i.e the one time in the game when a walk contributes to a win as much as a grand slam. In any other inning, the grand slam is the obvious choice.
Presumably the batter is aware this is the one time when a walk contributes as much as a grand slam. I’d be comfortable penalizing him for not being aware of that.
Because a batter hits a homerun, does not make him unaware of a situation. How would you feel if both batters had a 3-2 count, a fastball down the middle is thrown, one hitter fouls it off and then walks on the next pitch, the other hitter, hits the fastball over the fence.
I would feel that both batters contributed the same number of wins to their team by leveraging different skill sets. A guy who can reliably jack mistakes is preferred overall, but in this situation I would absolutely want the batter to prioritize getting on base over getting a hit. For someone like Fowler, for example, that might entail fouling off a meatball and taking a ball rather than swinging for the fences.
Here’s the thing: the batter had no hand in creating the situation at that time. What if it were the seventh inning with in the same scenario? The grand slam is more valuable.
The hitter didn’t decide to make it the ninth inning, he didn’t decide for his team to be tied by that time, he didn’t decide to load the bases right before his PA, etc… I don’t like making the walk more valuable than normal due to circumstances outside the batter’s control.
I don’t understand why you’re asking us these questions if the goal is to make a better metric. We aren’t professionals and you are. Maybe some of us are very knowledgeable but clearly many of us aren’t. Research isn’t a matter of taking popular votes and seeing what people think.
I agree. I mean, I voted in the poll, but I’m also very dumb.
From Tango’s first post in this series:
“Whereas Baseball Reference and FanGraphs and the others provide a consistent, systematic implementation, most fans have their own personal mish-mash of arbitrary, biased, and capricious combination of stats, which can change as their mood fits.”
I think the point of this exercise is to have a systematic way to capture a consensus of the “gut feel” part of what WAR is supposed to represent. Rather than just basing this on selected anecdotal evidence, Tango’s polling interested fans without a heavy prerequisite of expertise.
This is essentially the WAR equivalent of the Fan Scouting Report ratings, another Tango invention.
A lot of this comes down to what fans want WAR to represent. Let’s say you want to be able to use WAR to evaluate a award like MVP. If you feel like that award should go directly to the player that most helped his team win, then you probably wouldn’t care if said player drew a bases loaded walk in a tie game in the bottom of the ninth, or hit a homerun.
On the other hand, other fans, including myself want WAR to be a context neutral tool. If WAR is context neutral you can better evaluate which player was/is actual better. Basically, I want my WAR to reflect that Bryce Harper had a better offensive season than Anthony Rizzo, even though Anthony Rizzo had a greater influence impacting his team’s wins.
You an make the metric either way. I think which way the metric is calculated is based on what people want the metric to reflect.
John: if I feed you a WAR fish, that might satisfy you for a day, but what are you going to do tomorrow? Maybe you’ll blindly accept Fangraphs or blindly accept BR.com on the idea that those guys know what they are doing.
But, I’m trying to make you think about how you think about formulating and evaluating a WAR fish. I want you to have a reason to accept on or the other or both or neither.
I’m laying out all the considerations, or at least some key considerations, so you know why decisions are made as they are.
In EVERY QUESTION I have and will pose, there is no wrong answer.
Just as you can have a Philips screwdriver of different forms, or maybe a nail and a screw, you have to know why you are using the tool.
And I’m trying to figure out what kind of tool people are actually looking for. Maybe all we have so far are Philips screwdrivers, and somebody hasn’t made the hammer that you really need to use those nails of yours.
I want something which in the long run will work for evaluating how much a player contributes to his team, even if it isn’t what I think it should be now.
I voted for ‘same impact’ but actually I can easily imagine a case where it would seem to me that the walk should be valued MORE highly. If the count is 3-0 or the 3-ball pitch is clearly out of the strike zone, it would seem reasonable to penalize the batter for being so stupid as to swing at the pitch and take a chance on what happens in a situation where the win expectancy is very high if he takes it.
Assume in both cases it was a 3-0 count that preceded either the walk or the HR.
Well then I’m pretty sure I’d vote that the walk should be counted for more than the home run, which wasn’t even a choice. What kind of a stupid batter is going to swing at a 3-0 pitch in such a situation? Even if things happen to turn out well and he hits a home run, there’s no way that swinging in such a situation can be anything other than a bad decision which suggests the batter has terrible judgment,
Fair enough. How about you assume it’s 3-2 then.
Two questions:
1. What does “alot” mean? You’re asking us our opinions but we don’t have enough information to answer if we are considering options other than the two events are equal. Is alot 500x? 2x? I’ll treat it as “more” unless further clarification is provided.
2. The context of the pitch count matters here. Is it the same as yesterday, assume 3-0? If so, I would agree with other commenters then and say a walk may be more valuable than a HR. But if it is 0-0, I would swing the results in the other direction and make a HR more valuable. In that case, a HR is mostly within the batter’s control. It will take a pitcher throwing a hittable pitch. But other than that one event it is up to the batter. For a walk from 0-0, the pitcher has more control, as he will have to throw four balls. If the pitcher is (was) Clifton Phifer and throws only strikes, even Votto can’t walk.
Yes, it means “more”.
And yes assume 3-0. Regardless though you aren’t just evaluating the one pitch, but the whole plate appearance. IT’s not like the batter was gifted a 3-0 count to start the PA. He worked the count.
I’m so divided. The outcomes mean the same thing, but to me the HR takes more skill.
But then I wonder, does that skill make it more valuable?
To that I, personally, respond yes. So I voted for the HR to count for more. Hitting HR’s is valued higher than taking walks in this league.
You say it takes more skill, but not all home runs are created equal. A pitcher hanging a curveball and it going 10 feet over the wall really isn’t that difficult for the average major league player. Granted they do have to actually swing and make contact, but in some parks a pop up on a windy day is a home run.
Hitting a homerun does take more skill.
Regardless how is this an argument for using context dependent information for WAR?
He could also bury that CB 55 feet away for an easy ball four.
I’m just saying it’s worth more because look at the all or nothing home run hitters. They get playing time somehow. Not too many eye-only players stay around.
Talking about ballparks, that’s a completely different area to give value to. That isn’t really part of the question.
I think one’s answer to the question simply boils down to this scenario: Imagine you are standing 10 feet into the infield from third base staring at the runner on third. You can see nothing but him, you cannot hear anything, and you are locked on him the entire time. He is the only form of sensory information you have about the game. You watch him trot from third base to home to which you assume wins the game because his team crowds around him to celebrate. What is more important to you after witnessing it? Only that the team won the game? Or who won the team the game?
Correction: Or how the game was won and by who?
C’mon: two polls in a row with “alot.” A lot. A little. Abad mistake to make twice.
It was intentional. See my link in the other thread.
That makes it worse, not better. I’m not a prescriptivist by any stretch of the imgination–language rules should follow use and function–but the logic of your “improvement” is taking a common mistake, retroactively creating a very unlikely source of confusion, and forcing a rule around it.
“Lot” does functions the same as any other quantifier, so there is no compelling linguistic case for “alot.” It came into being because people with poor grammar/spelling were confused by the presence of other two word consolidations that actually have distinct meanings like “along” and made the leap. You can join the people making the mistake and the world won’t end, but you’ll willfully be making a mistake that detracts from your writing. And if you’re going to do it, at least have the decency to use it with all qualifiers (“abunch of people”) for maximum consistency.
*strike “does” from “does functions.” Being grammatical myself would be ag ood start.
This really highlights the challenge of dealing with such an abstract “stat” such as WAR. A BB and a HR are not equal events. Two players with 100 HR + BB: Player A with 50 HR and 50 BB and Player B with 10 HR and 90 BB… Player A is more impactful, more “skilled,” will contribute significantly more value over a season (more runs and more wins) and will make more money.
But in the particular instance described, the value to the team is the same. How do we reconcile this? I had always looked at WAR as the accumulation of a player’s individual accomplishments, weighting each event (PA, on base advancement or defensive play) based on the overall value of such an event (i.e., what is the relative weight to team success).
But that’s too simplistic maybe. The player in this scenario knows all he needs is a BB, not a HR. But it takes more skill to hit the HR. And being able to hit a HR in a high leverage situation seems possibly more of a skill, and more of a value overall (not just in an isolated instance), than being able to draw a BB… but maybe that’s just the classicist in me talking. I really don’t know. I am not sure I am sophisticated enough, or have the breadth of knowledge, to answer this question to my own satisfaction.
I think I am forced to choose making them equal, if only because it feels “right” to me to have the wins add up. After all, we analyze contracts on $ per win, so the control seems necessary and feels more like a balanced equation.
I’m not really sure why myself, but yesterday I voted for the home run counting for more than the walk, while here I voted for them to count the same. At this stage of the game, the most important thing is to win the game, regardless how that’s done. So the batter needs to have the ability to recognize which pitches to lay off (and thus draw the winning walk) or to crush (walk-off home run).
In other situations, the ability to hit home runs is more valuable than a routine walk, because it generates runs, which lead to wins, which are really nice. But in the scenario laid out here, what matters most is sealing that win. I don’t know; this is tough!
Like last post, I am not concerned with a specific context. I am concerned with how many runs are generated on average between a HR and a walk.