Statistical anomaly or new norm?

[rdivaldi]

If you gave the pitchers more strike zone to work with, it would change the dynamics at the plate and open the game. You would have fewer walks, more in-play hits and, once the hitters figured out the new landscape, strikeouts would be less frequent because they couldn't sit on mistakes....Give pitchers more room in the strike zone, and hitters will figure out that they can't sit on pitches in their zone.

I've seen this argument before, but I think this would be a recipe for disaster. If you expand the strike zone, pitchers would likely take advantage of the newly defined edges and nibble even further away from the middle of the plate. This would equate to even lower batting averages and more strikeouts. There's no way to get rid of the current home run surge via traditional rules, the only way to fix it is to deaden the ball or force teams to push back the outfield fences. Since we all know teams are not going to push back the fences, here comes a dead ball controversy...

[ChiTownTrojan]

That logic is flawed. Why would a home run count any differently if the team is tied or down 2, 3, 4, 5, 6, etc. runs? What if a player hits a home run to put his team down 3, then his teammates continue the rally to win the game? Does that home run count more? A run is a run is a run. Why would it penalize hitters for striking out? An out is an out is an out. Just because you don't like something doesn't make it worse than other outcomes.

There are certainly cases where a run is more important to winning a game than other situations. If your team is down one in the bottom of the 9th with two outs, and you hit a 2-run HR, that home run was the difference between winning and losing. It could be argued that a better definition of WAR would be to give that player a full point of WAR for that HR (rather, a point minus whatever a replacement player would be worth in that situation). This actually might be possible to implement - there are new stats that calculate the "win probability" for any given situation, and you can determine exactly how each at bat changed a team's probability of winning a game. Add up how the win probability changes for every at bat in the season and you might get something that looks a lot like oWAR. Call it soWAR for "situational oWAR" or something like that. I'm sure there are flaws with this method (do we really want to overvalue late inning at bats to that extent?), but I'd still be curious what would come out of it.

I think the argument against this approach, which TDog referenced, is that WAR purposely doesn't take into account situational hitting, for two reasons: (1) a player is not responsible for the situations that he comes up to bat, so he shouldn't get extra credit for being in those situations, and (2) situational hitting numbers are not reliable due to small sample size, and a better estimate of a player's true situational hitting ability would just be to look at their total stats. There is some evidence to support this, but some outliers (like Abreu) that really do look like they change their approaches to adjust to the situation.

Part of the issue is whether WAR should be considered a stat that reflects how a player actually performed, or whether it should reflect a player's true underlying ability. I would argue that WAR should reflect what actually happened on the field. In many ways, it already does that. For example, this year, in the small 60-game sample, Jose Abreu made a bunch of outstanding defensive plays. I don't think he is actually one of the best defensive 1Bs in the league, but he made those plays this season, and they helped the team win games. His defensive WAR reflects the fact that he made those plays, not what we think his true ability is or what he will do defensively in the future. Conversely, Madrigal made lots of mistakes this season defensively and on the base paths, and he is rightly penalized for those mistakes in the WAR calculation, even if most of us expect these issues to improve and mostly be a result of the small sample size. Following this logic, I think WAR would be improved by incorporating situational hitting into its calculation, which would better reflect how a player actually contributed to wins and losses, rather than that player's true abilities. Or at least you can calculate two different versions of WAR, which describe two different things (actual W/L and hypothetical W/L). Actual W/L would be better used to determine an MVP, for example, whereas the hypothetical W/L might be better if you are a GM looking to build a team for the following year.

[FourGoldGloves]

This discussion makes me wonder which is more real and more compelling, the game of baseball or the game of WAR. It reminds me of the short story by the Argentine writer Jorge Luis Borges, "Tlön Uqbar, Orbis Tertius", about an imaginary world described in more detail than the real world so that it eventually overcomes the real world.

[voodoochile]

There are certainly cases where a run is more important to winning a game than other situations. If your team is down one in the bottom of the 9th with two outs, and you hit a 2-run HR, that home run was the difference between winning and losing. It could be argued that a better definition of WAR would be to give that player a full point of WAR for that HR (rather, a point minus whatever a replacement player would be worth in that situation). This actually might be possible to implement - there are new stats that calculate the "win probability" for any given situation, and you can determine exactly how each at bat changed a team's probability of winning a game. Add up how the win probability changes for every at bat in the season and you might get something that looks a lot like oWAR. Call it soWAR for "situational oWAR" or something like that. I'm sure there are flaws with this method (do we really want to overvalue late inning at bats to that extent?), but I'd still be curious what would come out of it.

What about all the pitchers who worked to keep the game within one run? What about the other players who scored or drove in a run to get the game within 1? What about the guy who reached base before the 2-run bomb? Does the opposing pitcher lose a full point of WAR because he yielded the HR? What if the ball deflected off a leaping OF glove - does that player lose a point? What if the ball eluded a diving RF and rolled all the way to the wall allowing an inside the park HR? Does that OF lose a point?

It's too simple to give a single player a full point for a single play. Any game that is decided by one run had many people contributing to the win and the loss.

[ChiTownTrojan]

What about all the pitchers who worked to keep the game within one run? What about the other players who scored or drove in a run to get the game within 1? What about the guy who reached base before the 2-run bomb? Does the opposing pitcher lose a full point of WAR because he yielded the HR? What if the ball deflected off a leaping OF glove - does that player lose a point? What if the ball eluded a diving RF and rolled all the way to the wall allowing an inside the park HR? Does that OF lose a point?

It's too simple to give a single player a full point for a single play. Any game that is decided by one run had many people contributing to the win and the loss.

Like I said, I'm sure there are flaws with the system, but I'd still be curious what came out of it. Also, I was talking about using it to calculate oWAR (offensive WAR), so most of the examples you gave are not relevant. It's not clear how errors would be taken into account, because no I don't think a hitter should get credit for winning the game if he hits a ground ball that is thrown into the stands by the SS and ends up scoring two runs. But on a straight up hit (or out), with no significant defensive contributions to how the play went, I think a good estimate of how that play factored into whether the game was won is the change in win probability before/after the play.

[rdivaldi]

It's too simple to give a single player a full point for a single play. Any game that is decided by one run had many people contributing to the win and the loss.

Exactly. Recency bias is in full effect with some of these comments. Forget the first 8 innings and focus on the 9th.

[Mohoney]

Straw-man argument. What's "icky" is everybody swinging for home runs all the time, and the resulting huge increase in strikeouts. There have been more strikeouts than hits for the past couple of seasons, something that had NEVER happened before. Again this season, 15,586 strikeouts, 14,439 hits. In roughly a third of a regular season. Watching strikeouts is "icky." It's boring. Putting the damn ball in play gives you something to watch, some action. "Pace of play." Players like Abreu can hit monster home runs, but also hit a clean liner to right to score a runner when it's needed. There aren't many like him.

Teams can be behind by 3 or 4 runs in the ninth. They need baserunners. But they all go up there trying to hit solo homers, and strikeout, strikeout, strikeout. That is what is "icky."

This actually proves the point. More runs are being scored despite the increase in “icky” strikeouts.

Without a drastic recalibration of the value of a single relative to the value of a home run, the situation is not going to change.

Personally, I consider ground balls to be every bit as unappealing, mundane, boring, etc. as strikeouts. However, I would never erroneously claim that ground balls are less conducive to run production than strikeouts.

[ChiTownTrojan]

Exactly. Recency bias is in full effect with some of these comments. Forget the first 8 innings and focus on the 9th.

Apparently my idea isn't that crazy, because FanGraphs already calculates it:

https://library.fangraphs.com/misc/wpa/
https://www.fangraphs.com/leaders.as...2-31&sort=14,d

Yes, at bats in the late innings are worth more according to this metric than those in the early innings. That's the point.

[voodoochile]

Apparently my idea isn't that crazy, because FanGraphs already calculates it:

https://library.fangraphs.com/misc/wpa/
https://www.fangraphs.com/leaders.as...2-31&sort=14,d

Yes, at bats in the late innings are worth more according to this metric than those in the early innings. That's the point.

I'm not arguing that a 2-run bottom of the 9th blast in a 1-run game isn't worth more points, but a full point seems too much and would skew the metric dramatically.

[TDog]

This discussion makes me wonder which is more real and more compelling, the game of baseball or the game of WAR. It reminds me of the short story by the Argentine writer Jorge Luis Borges, "Tlön Uqbar, Orbis Tertius", about an imaginary world described in more detail than the real world so that it eventually overcomes the real world.

Borges spoke at my college when his Nobel Prize near the end of his brilliant career seemed imminent and, recalling whimsically for a moment what the world was like more than four decades ago, we believed we would change the world with out writing (spoiler alert ...). But magical realism is kept alive in baseball. Or is it magical realism keeping baseball alive? Or is the question more important than the answer?

From a field of dreams to cold, hard 2020 where baseball is dying. The commissioner hasn't used those words, but his rule changes and proposals for rule changes reflect such desperation. Last season, there were even people throwing around the idea of shortening the game to seven innings, and this year, as noted in the thread, you have the three-hitter rule for relievers because baseball is obviously wasting our time.

Some won't understand my point. For that matter, the commissioner doesn't have enough love for baseball to understand the difference between between hitting and offense, between the pace of the game and the length of the game. I work under the handicap of having seen baseball when it was a hell of a lot more fun, despite following the White Sox, who weren't necessarily fun. Some days were more fun than others. On a Sunday afternoon about six weeks before my 13th birthday, I saw the White Sox score 22 runs in a game in Boston. The Red Sox only scored 13. Bill Melton led off the third with a home run. The other 23 White Sox hits stayed in Fenway Park, as Ken Harrelson might have said, but the Red Sox traded him to Cleveland a year earlier. Only 18 RBIs as Carl Yastrzemski, in center, dropped a fly ball clearing the loaded bases. It was Yastrzemski's second error in the top of the first Misplaying Luis Aparicio's RBI single after Walt Williams' leadoff double was another factor contributing to the six-run first that knocked out Gary Peters. Carlos May later scored on a Ken Brett wild pitch -- from third with one out with Melton at the plate. Six years later, in a 1-0 11-inning White Sox win in Anaheim, Melton, hitting cleanup for the Angels, would get the only legitimate hit against the White Sox version of Ken Brett with two outs in the bottom of the 10th. George's big brother lost his no-hit bid with two outs in the ninth when Jorge Orta misplayed a ground ball at third but didn't get charged with the error..

Baseball is a team sport, but players' individual actions are so well defined, days -- weeks, years, even decades later -- that the data dump can be immense. Given 21st century computers, you don't have to be an MIT professor to work out theories of percentage baseball. It's even tempting to develop the equivalency of a unified field theory for comparing the relative values of player. I find it just as elusive in baseball as it is in physics, but others have signed up to believe in WAR. Yastrzemski had a great offensive year in 1970. He hit 40 home runs and lost the batting title to Alex Johnson on the last day of the season. That night in Anaheim after the Red Sox were done,,Alex Johnson passed him up with a ground ball to third, misplayed by Melton and ruled a hit. Jay Johnstone pinch ran and Johnson celebrated his batting title. But in 1970, the Red Sox were never in contention, finishing third in the East. Still, Yastrzemski recorded a WAR of 9.1. In his defense, he only played a few games in center. But Dick Allen two years later -- carrying the White Sox in the unlikeliest of division races, falling off the pace of the Reggie Jackson-Catfish Hunter A's dynasty in September -- only recorded a WAR of 8.6. I'm not willing to look at the two players' lines and say Yastrzemski had the better season. I'm certainly not going to concede Yastrzemski's 1970 value based on their respective WARs, and I would be disappointed if anyone here did so.

You can isolate what a player does, but it's a mistake to isolate him from the circumstances in which he does it. Madison Bumgarner's appearance in Game 7 of the 2014 World Series wasn't just five innings of scoreless relief. Because we're talking here about a ;regular season (pseudo) statistic, fill in your favorite regular season performance -- perhaps John Danks' eight-inning, two-hit scoreless start with never more than a one-run lead in the Black Out Game. But I believe a unified field theory of MLB player value is neither possible nor necessary and don't understand how it could be considered remotely fair to count a tie-breaking late-inning home run (or Buckner-esque game ending error) as one would judge the same play near the end of a blowout game. How you weight a player's successes and failures is at the heart of WAR's bias. You can isolate what a player does, but what he does is in the context of his team, who his is surrounded by in the lineup and on defense, what situations he has to face for the good of the team. With WAR you end up with numbers that inspire Web pages arguing that Ray Schalk should be removed from the Hall of Fame. I don't believe anyone alive can remember seeing him play, but such sites claim a higher authority than the contemporary accounts of Schalk's dominating presence behind the plate, even being the first catcher ever to tag a runner out at every base. And he accomplished the feat at second more than once.

I have no doubt that people here get as much enjoyment out of WAR and statistical analysis as they do watching baseball, or that simply winning is fun regardless of how much fun there is in the game. Baseball doesn't seem content with what has become of the game, hence the three-batter rule for relievers, which does nothing to solve the obviously out-of-joint ratio of hits to strikeouts, not to mention the pace of the game. Unlike the 24-second clock, the NBA rule change after the 1954 season that forced players to shoot the dam ball, no MLB rule change has forced hitters to swing the damn bat.

When I suggest a bigger strike zone, I'm not suggesting a radical change, akin in the aforementioned reliever rule. I'm talking about opening up the game by rolling the strike zone back to where it used to be, if not when the Yankees hit .307 as a team, at least to when Dick Allen hit .307 for the White Sox.

[ChiTownTrojan]

Where does this idea come from that baseball is dying? For a dying sport, a lot of people sure are watching it and paying attention to it. Is there an objective measure that says that baseball is dying, other than number of games per season that TDog is watching?

I did hear attendance was way down this year...

[Frater Perdurabo]

Where does this idea come from that baseball is dying? For a dying sport, a lot of people sure are watching it and paying attention to it. Is there an objective measure that says that baseball is dying, other than number of games per season that TDog is watching?

I did hear attendance was way down this year...

It’s “dying” in the minds of seasoned fans who don’t like the way the game has evolved.

[FourGoldGloves]

But magical realism is kept alive in baseball. Or is it magical realism keeping baseball alive? Or is the question more important than the answer?

I think how much richer the magical realism would be if it were describing the old game, rather that the current Home Run Derby, where the conclusions of the statistical analysis are more obvious and only validated by their obviousness. E.g. a home run scores the baserunner no matter where he starts. So don't take chances on the bases, and for heavens' sake don't bunt. duh.

[ChiTownTrojan]

It’s “dying” in the minds of seasoned fans who don’t like the way the game has evolved.

I spend a lot of time at parks these days, with two young kids participating in a total of 5 sports leagues. Almost every time I go I see baseball being played, either a game or a practice. My 6-year old is in his second year of baseball, he's on a team with a bunch of his friends, and they all love it. He also got really into the Sox this season, most of the games were past his bed time (we're on the east coast) but he would wake up wanting me to tell him what happened in the game and watch the highlights on my phone. When there were earlier games he would get really excited, and we would watch together as much as we could. His favorite players are Luis Robert and Tim Anderson, because as he says, "they're really fast and I'm fast too."

From my perspective the game is doing just fine. I would prefer there to be fewer strikeouts and more contact, but it hasn't impacted my interest in the game at all.

[TDog]

Where does this idea come from that baseball is dying? For a dying sport, a lot of people sure are watching it and paying attention to it. Is there an objective measure that says that baseball is dying, other than number of games per season that TDog is watching?

I did hear attendance was way down this year...

The idea that baseball is dying comes from the commissioner.