WAR Games
July 21, 2020 11:04 AM   Subscribe

In such a complicated numerical landscape as a 60-game season, advanced stats might approach their greatest hurdle—but also, counterintuitively, their greatest opportunity to aid understanding above and beyond traditional stats. from The Ultimate Baseball Stat Confronts Its Ultimate Test [The Ringer] posted by chavenet (31 comments total) 7 users marked this as a favorite
 
I know the reasons for it, but having the off-season last until July and then seeing that the Nationals have the second day of the season off feels like an unbearable indignity.
posted by Bulgaroktonos at 12:10 PM on July 21, 2020 [1 favorite]


New situations require new stats. What about games before entire team becomes hospitalized? Wins above trainer dying? Bases per infection?
posted by fifteen schnitzengruben is my limit at 12:24 PM on July 21, 2020 [14 favorites]


WAR Games

I would have also accepted WAR: What is it good for?
posted by zamboni at 12:43 PM on July 21, 2020 [17 favorites]


free fantasy (in multiple senses) baseball team name/concept: WARforged
posted by dismas at 12:54 PM on July 21, 2020 [1 favorite]


I've literally never watched an entire professional baseball game.

I haven't played any kind of baseball since I was 6 years old.

But, goddam, do I like reading about this baseball nerd shit!
posted by Reyturner at 1:14 PM on July 21, 2020 [5 favorites]


WAR-riors come out and play-eee-ayyyy
posted by dismas at 1:34 PM on July 21, 2020 [4 favorites]


New situations require new stats. What about games before entire team becomes hospitalized? Wins above trainer dying? Bases per infection?

WAVE, Wins Above Viral Exposure.
posted by madajb at 2:13 PM on July 21, 2020 [14 favorites]


I should know better than to argue with random bloggers but from the "WAR is a dumb stat" link:

One last point that I will touch on about WAR, if a players replacement is not as good as them, their WAR will be significantly better as well, hence “wins above replacement.”

That's... not how WAR works, is it? That's not how it worked back when I was paying attention to this, anyway. The variation in the meaning of "above replacement" is mostly just the position adjustment (that the writer is also criticizing) and is meant to be relative to the caliber of position player that any team could call up from minors, not your actual teammates. The idea is that on average a catcher who can hit, say, is worth extra because most catchers can't.

Also while the defensive stuff is definitely a dark art, his own chart shows Trout's offensive numbers to be better, not worse - is this the last person to remain unaware that walks are good?
posted by atoxyl at 6:10 AM on July 22, 2020 [4 favorites]


People don't argue against what WAR is, they argue against what they suppose it is from what Joe Morgan told them or what they half listened to on sports talk radio.
posted by DirtyOldTown at 6:20 AM on July 22, 2020 [4 favorites]


I get the impression that to be a baseball fan, one has to be a numbers geek. I’m sure fans exist who watch the activity on the field for fun and don’t pore over the numbers to make what happened meaningful. But even before Billy Bean and Sabremetrics, Baseball had a reputation of being math-heavy in a way more action-and-collision-oriented sports are not.

Is baseball interesting without the math, or is that a big appeal to the fans? Do people gravitate to baseball fandom for the math?
posted by Pirate-Bartender-Zombie-Monkey at 6:52 AM on July 22, 2020 [1 favorite]


It's not the math itself that many baseball fans are really into so much as the way the math more or less allows people to compare players across baseball history in a way that other sports don't. It gives a much greater sense of the game being a continuum, where the current players aren't just playing against this season's opponents to win games, but are becoming a part of a history that stretches back to before the Civil War. WAR as a stat provides some evidence of how players might measure up to others from earlier eras in a semi-direct sense as the more static nature of the game, a pitcher faces a batter who tries to get on base has remained roughly the same, allowing for some changes in rules, over time.

Given the length of the seasons and how few teams actually make the playoffs, the enjoyment of the game for many is in seeing the stats accumulate and being able to argue that, say, Mike Trout might be a better player than Mickey Mantle, or just have a basis for supposition as to how Walter "Big Train" Johnson might match up against Randy "Big Unit" Johnson and on and on. Baseball is as much about its past as its present in that way, and in the same sense allows for more ample speculation about the next young groups of rookies and prospects based on their stats as well.
posted by gusottertrout at 7:10 AM on July 22, 2020 [9 favorites]


Is baseball interesting without the math, or is that a big appeal to the fans? Do people gravitate to baseball fandom for the math?

To be honest, I always found that the numbers get in the way of my enjoying baseball. I just like to watch a game unfold and evolve, enjoy the pitcher/batter duels, see a well-turned double play, etc. It just never feels like the numbers nerds actually enjoy the game.
posted by Thorzdad at 7:50 AM on July 22, 2020 [3 favorites]


A key reason I love baseball is because it is an object lesson in how most things that look like magic are actually math.
posted by DirtyOldTown at 8:27 AM on July 22, 2020 [5 favorites]


The war against WAR is just the sabermetric version of the perfect being the enemy of the good. Nobody who's spent time studying baseball analytics can make a convincing case that conventional metrics like batting average, runs batted in, and fielding errors do a better job of capturing a player's overall value than WAR, yet the first sign of any discrepancy between WAR values and someone's subjective evaluation of who the better player is can lead them to conclude that WAR is bunk.

ProTip: when there are 3 prominent versions of a statistic that all disagree with each other, you're not exactly breaking new ground when you note discrepancies.
posted by tonycpsu at 8:38 AM on July 22, 2020 [5 favorites]


look some of us just have a vorped perspective on the games
posted by Huffy Puffy at 9:18 AM on July 22, 2020 [4 favorites]


look some of us just have a vorped perspective on the games

I am imagining you dropping that line like David Caruso in CSI, only instead of whipping the shades off, they're the ones outfielders wear and you're flipping them up.

too inside baseball?
posted by DirtyOldTown at 9:27 AM on July 22, 2020 [2 favorites]


Do people gravitate to baseball fandom for the math?

Michael Lewis drew me in with the math. Jon Bois kept me in by demonstrating how baseball being silly and arbitrary is good, actually.
posted by Reyturner at 10:35 AM on July 22, 2020 [2 favorites]


now I wanna yell “yeaaaaahhhhh!” and bang on a trash can.
posted by Huffy Puffy at 10:58 AM on July 22, 2020 [2 favorites]


I mean it is a feature of baseball to me that you don't actually have to watch all the games - it's well-suited to summary. Not that I don't like watching some of the games, especially in person.
posted by atoxyl at 11:22 AM on July 22, 2020 [5 favorites]


I was never a baseball player, though, which might be a difference from some of the folks who don't care to hear about the numbers.
posted by atoxyl at 11:26 AM on July 22, 2020 [2 favorites]


Michael Lewis drew me in with the math. Jon Bois kept me in by demonstrating how baseball being silly and arbitrary is good, actually.

That's something of the trap that WAR, or the desire to completely quantify athletic achievement as is happening in all sports, leads to. The want to fully capture and compare the achievements of competitors across different teams, eras, and circumstance and through that to then be able to predict who will succeed and fail, get injured or stay healthy and so on, is a completely understandable urge. It has both the excitement of discovery in the attempt to be the person to make sense of all the statistics that are kept around sports and allows for a like feeling of contribution to the sport one loves.

But at the same time that desire to quantify athletics, like similar desires to capture "art" and so many other intrinsically human activities and render them into data to be played with to "better" the ability of teams and fans to pin down exact values of every action and career, works against the very nature of sports, the arts, and so many other things people take pleasure in. I mean a predictable sport is no sport at all and the closer statistics come to "leveling the playing field" the less interesting and exciting the sports become.

At first it just takes away the debates about who was best, how to compare player X to player Y on different teams or in different contexts, but then works to making players and teams "more efficient", driving teams to play the same way to best work the rules of the sport, to push athletes to adopt the same techniques for the most efficient method of play and create more predictable results, and then using those same techniques to assess the next generations of athletes who best exemplify the "efficiencies" desired by teams. Signing players becomes less and less arbitrary and more predictable and who wins and loses starts to become ever less interesting as well. Sure, there would still be the individuals of exceptional ability showing what they can do, but the whole becomes more routinized, designed to eke out little advantages at the margins, assisted by yet more data crunching to determine just where those margins might be, and ever less accessibly "human" in the fuller sense of the term as the silly and odd gets the boot for the "pleasures" of fucking data.
posted by gusottertrout at 12:46 PM on July 22, 2020 [2 favorites]


a predictable sport is no sport at all

Baseball stats are easily the least likely to lead to predictability: the game embraces randomness to a degree far, far beyond most sports. You can rarely predict what is going to happen, but you can sure as hell aspire to count it. This is why baseball stats are so hypnotic.
posted by DirtyOldTown at 1:54 PM on July 22, 2020 [3 favorites]


I have to disagree. Have you seen all the data they keep at Statcast and what it's being used for? They track defensive positioning, batter swing planes, loft, how hard the ball is hit and the angle it is launched in order to refine those things to the closest ideal, either in the players chosen or in getting players to adopt similar strategies.

They track the arm slot of pitchers, rotation, movement and MPH of every pitch and likewise try to get pitchers to adopt similar arm slots and motions for maximum effect, while developing the best pitch mixes for them thanks to computer tracking, and then mix the pitchers themselves to maximize use against individual batters by identifying patterns, which isn't just about banging on cans and moving the defense around to the most likely areas the batters will hit the ball to, tracking the fielders by camera for how quickly they get a jump on the ball in play, the optimal route to take to field the ball and how fast they reach it from their starting position among many other things. And that's only the public stuff, with all but the most stodgy of teams locked in an arms race for better data and analysts to refine their approaches even further.


Strikeouts, HRs and lots of exciting relief pitcher changes that drag the games on for hours was the result, which the fans enjoyed so much MLB had to change the rules to try and stay ahead of the data. Teams chase the same info and come to much the same results, but not entirely as their are still dinosaur teams making bad trades and not caught up with the others and the whole area is still in the early stages of being strip mined for more and more information, which does virtually nothing for the the pleasure of viewing a game but becomes a necessity in order to stay competitive, and as the info is further refined the edge it gives becomes lessened as more teams adopt similar "optimal" strategies to reduce randomness as much as possible in every aspect of the game.
posted by gusottertrout at 2:27 PM on July 22, 2020 [1 favorite]


Oh, and I suppose it's almost needless to add that the owners use this same kind of data as a way to best screw over players as well, but I'll mention it all the same for those who don't follow the sport.
posted by gusottertrout at 2:31 PM on July 22, 2020 [1 favorite]


That data can tell you probabilities but it can't get close to predicting the odds of even most micro-events within the sport, let alone the outcomes of single games. Data gets series wrong all the time. Even entire seasons.

The competitive advantage good data gets you is being able to shift a positive outcome from occuring say 25% of the time to say 26.1% of the time. That small increase has a cumulative value based on the sheer number of games, but it does essentially nothing toward predicting the outcome of any given situation.

Joey Votto is a great hitter, who has an amazing career OBP of .575 against Kyle Hendricks. But that means he is out 42.5% of the time. All of that data isn't going to clear 42.5%. It might might get his success rate up to 58, 59% if he isn't maxed out already. (Maybe he got to that 57.5% by data!) But meanwhile, Hendricks' team is compiling data, too. And Hendricks could be getting better. And Votto might have eaten a lunch that upset his stomach. Or he's getting older. And Hendricks might get a new catcher who calls a better sequence for Votto. And on and on.

In baseball, the very best teams lose 40% of the time. The best pitchers routinely allow enough runs to be beaten. The worst hitters crank a home run here and there. It's got randomness baked right in. No stat will ever change that.
posted by DirtyOldTown at 3:33 PM on July 22, 2020 [2 favorites]


At first it just takes away the debates about who was best, how to compare player X to player Y on different teams or in different contexts

I almost feel like the opposite has happened - that there are more of these conversations than ever - and also that these conversations are pretty boring but on both counts that might just be the internet doing things to death.

I think you probably can make a case that sabermetrics has made baseball strategy and tactics a bit less interesting, though, by showing that a lot of the more risk-taking plays don't work very well and that playing to either hit a home run, walk, or strike out does. Not to mention that one reason those outcomes are favored is that they are the easiest to quantify. That all plays into things that were happening to baseball anyway but it's not great for variety.
posted by atoxyl at 8:22 PM on July 22, 2020 [1 favorite]


In baseball, the very best teams lose 40% of the time. The best pitchers routinely allow enough runs to be beaten. The worst hitters crank a home run here and there. It's got randomness baked right in. No stat will ever change that.

The goal, quite simply, is Taylorization of the sport to maximize efficiency and, not coincidentally, to give greatest benefit to the owners of the teams by removing as much of the more random human element as possible from the sport. That all random events can't be controlled for isn't as much the point as the desire to erode as much of that "messy" element as possible so teams can better control outcomes. This is still at the early stages of the process, not the end result, so it will continue towards ever greater routinization as that is seen as providing owners the greatest benefit even as it eats away at many of the elements that drew fans to the sport in the first place.

Tampa Bay is the model of this attempt, where players are shuttled in and out to fit formulas for best result in both play and contract control. As the most extreme outlier at the moment, they're a fun team to root for in that sense as they are innovating new ways to get more from their "pieces" in every aspect of the sport, treating players like interchangeable machine parts that they fund through novel accounting, to be discarded the moment the cost benefit analysis shifts. It's good for the team, bad for the players and creates a narrower window for fan engagement. It's no wonder major league baseball is losing so much of its fan base with the direction its been heading and I can't see that changing any time soon. Everything being done makes the sport a more specialized interest and makes it more difficult for the average fan to engage with over the way decisions are made. If it wasn't for fantasy baseball and the monetization of that aspect, things would even be worse.
posted by gusottertrout at 10:48 PM on July 22, 2020 [1 favorite]


The reason I’m so fascinated by the analytics of baseball is that I did an FPP a while back where I learned that Australian Rules Football is one of the most analytics-intensive sports out there. But it’s also an action packed competition that you pay attention to.

The sedate, 19th century pace of Baseball fascinates me in 2020. I do think it has been eclipsed as “America’s Passtime” by other games, and of retains that title out of generational inertia. @Atoxyl described it as a game well suited ti summary. There’s an X-Files episode about baseball where Mulder talks about reading box scores and being able to recreate a game in his head.

Nobody sits down and reimagines an Ozzie Rules match, even if they could read the transcript of that guy with the binocs giving the play-by-play. Different game, different experience. ARF stats feed the “fantasy” leagues, but analyzing some player’s “pressure” stats this season is fully different from looking at what a batter did during the 4th inning of game 2.

I myself am not a numbers guy. I’m here for the game in front of me, the score on the board. And no disrespect intended, I am fascinated by sports-numbers-folk, because it’s an exercise in trying to understand someone else’s joy that I don’t partake in.
posted by Pirate-Bartender-Zombie-Monkey at 9:45 AM on July 23, 2020


From the birdsite:
Washington Nationals star Juan Soto has tested positive for COVID-19 and will miss tonight's season opener against the New York Yankees, sources tell ESPN.
Stop. Just stop. Before you get people killed.
posted by fifteen schnitzengruben is my limit at 11:39 AM on July 23, 2020 [3 favorites]


Next you guys will be telling me that Giancarlo Stanton won’t be hitting 60 homers this year.
posted by Huffy Puffy at 4:20 PM on July 23, 2020 [1 favorite]


Game 2, and I might be wrong. Stanton might hit more than 60.
posted by Huffy Puffy at 5:27 PM on July 25, 2020


« Older I’ve never been more certain that Daisy is...   |   Beware not the Enemy from 'Without' but the Enemy... Newer »


This thread has been archived and is closed to new comments