Twins 8, Red Sox 6: Early Deficit Erased in Scrappy Fenway Win

I love this game called baseball.

38 minutes ago, JD-TWINS said:

I know they score runs……… 9 - 10 - 8 - 8 - 13 - 8 - 11 - 11 - 9 - 8  ……… 9.8 runs/game over 10 games ……. skews the total run output as it relates to being competitive regularly.

They have scored 2 or less eleven times through 51 games……… 3 or less 19 times. I get that “this is baseball” but there have been too many good pitching performances washed away due to lack of run support.

They have just sent down two guys (regulars) that had terrible, terrible offensive stats. They have James Outman - Caratini - Bell - still on roster with negative WAR numbers. Lee & Clemens & Keaschall have all been under 0.00 WAR for big parts of the season.

Jeffers - Buxton - Martin - Larnach have supported the stats in a big way to date.

To be “competitive” the offense needs to play better day to day.

The basic statistical concept of "mean" (i.e. "average") predominates the rate stats on every site, but I don't know of good sources for the next-level concept of "variance," which is the issue that I've long felt plagues the Minnesota offense.  It would take massive research, I think, to track down evidence to support my contention that, to a greater extent than most good teams, we feast when the opponent on the mound is subpar but regularly starve when someone of major league talent is on the mound.  Maybe such a study would convince me that my supposition is wrong...

35 minutes ago, JD-TWINS said:

I would suspect that the lack of success v. “good” pitchers and success against lesser opponents is pretty typical across baseball.

I remember a stretch where TWINS beat (and scored 5 plus in each instance) Skubal - Valdez - Gray - Crochet ……, so, ………….

Yep, the concept I am trying to get at is really difficult (for me anyway) to define.  "We scored 6 runs off their bullpen - they must have all been AAAA callups."  "We beat Skubal - we beat Crochet - they must have not been on their game that day."  These could be legitimate explanations - or just typical BS after the fact.

Maybe some important factors are just unknowable.  Or maybe Big Data techniques have ways of teasing out the meaning from scads of conflicting data.  I tend to believe that if an effect is important, then evidence for it can be found if you look at the evidence correctly.  (You can also find evidence for your supposition by looking incorrectly.)

It's a bit like, we all believe a .300 hitter is better than a .260 hitter, but when you come down to individual games, anything can happen (and often does).  Maybe that .260 hitter has just as many multi-hit games as the .300 guy, but he suffers more o-fers while the .300 guy plugs along with more 1-for-4 days.  That kind of variance is completely missed in season-ending averages, but might affect how you value the two players.

I spent my career in a small corner of analytics, but I never had the technical chops to tackle a statistically-oriented question involving variance for sets of data (like in baseball) that are a bit unruly to start with.  Uniformity is lacking - some days you get 3 plate appearances in a game, some days you get 5, and how do you do a "variance" study of how many hitless days you have, in that environment?

55 minutes ago, Linus said:

I agree with this. Maybe a simpler way to assess an offense is what percentage of the time do they score 3 or less runs (pick whatever number you feel assured a loss). Essentially how many times do we lose games because the offense doesn’t show up. I think this is where the Twins are suspect. Of course I’m not going to do the research so I will just assume I am correct😀

A bit of spot-checking along these lines a few days ago didn't turn up the pattern I was hoping to be able to demonstrate to others in support of my view, so I didn't post about it.  😀

Man if your Martin who doesn't hit many home runs.  Has to be pretty special to hit one over the green monster.  He's having a nice year so far.

2 hours ago, USAFChief said:

For much of the past century, since the deadball era, the magic number in MLB is 5. Score exactly 5 runs/game, you'll win more than lose. Score 4, you'll lose more than you win.

There's been a couple short exceptions. The late 90's, early 2000s saw that number reach 6 in several seasons.

Occasionally, particularly through the 60s and 70s, the number was 4.

I know this doesn't address the issue at hand--variability--but I've always found it interesting.

If you could somehow put up 810 runs, exactly 5 per game, you likely have one of the best regular season teams in baseball. 

It's a valuable observation and there are many potential tangents to take in response.  I like to test theories by looking for exceptions, and for example this season the Nationals are on a pace to score 875 runs, yet (as of this morning) are below .500 because they give up runs at an even higher pace.  However they're merely the exception that proves the rule because yes, a typical pitching staff and the defense behind them will make 5 runs stand up more often than not.  The context of a given season matters, as you point out, so that this year and the past two years the average runs scored has been about 4.4 per game but as recently as 2023 it was 4.6 so then your 5 runs gets you less of an edge than right now - and 2019 was even more out of kilter at 4.8.  Long ago, Bill James popularized an empirical observation that every 10 additional runs scored during a season (and conversely every 10 runs saved) works out to about 1 additional win - a ratio that at first blush sounds kind of ludicrous since ten key runs here or there could net you several wins not just one, but the distribution of runs across a season doesn't work that neatly.  Those are some random thoughts, and I could go further and further down this rabbit hole, but my main reaction to your message is simply "yup."

Still, I want to steer this back to variability, and your 810-run rule of thumb makes for an entertaining thought experiment.  Last year two teams scored 810, the Yankees and the Dodgers, and everyone envies their win totals.  As you say, if you score 5 runs a game in today's environment, you set yourself up for a successful season.  Nobody scores exactly the same number each game, but if you did and it was 5, you'd do well.  However, if 810 for the season is the target, then there are other ways it could happen.  Suppose you scored 10 in half your games but got shut out (even when you go into extra innings, let's say) in the other half; you'd still score 810, but at best you'd finish .500, and since most teams, even good ones, give up 11+ runs a handful of times in a season, you'd be odds-on to finish something like 78-84.  Let's make the variability even wackier: suppose you score 15 runs in a third of your games and get shut out in two thirds.  That still brings you to 810 runs, and assuming your pitching/defense never completely blows up, you are destined for a 54-108 debacle of a season.

That's a thought experiment and can never happen - but neither does a constant 5 run output across an entire season either.  The idea in the above paragraph is to demonstrate that variability could be really, REALLY important - the greater the variability the greater the potential effect - but as far as I know it's not really studied at all.  And as I said earlier, someone more schooled in industrial-grade statistical analysis than myself would need to be called on to even begin the process of defining what to look for.  Because batters, pitchers, and managers intentionally do things for competitive reasons that skew the uniformity of the data available for us to work with - the manager chooses relievers differently when the score is 5-4 than when it's 8-1, as a trivial example, and pitchers pitch differently to Matt Wallner when the game is not close and/or there is no one on base, for another.