The Time Has Come to Place Byron Buxton on the Injured List

[howeda7]

On 6/6/2022 at 6:40 PM, USAFChief said:

Do we know that an IL stint will fix the knee? 

 

If it will, then IL him immediately. He's not the same player.

 

If it won't, and I don't think it will or they'd have done it already, then we're getting the 100 games of Buxton they told us we were getting, and we're not getting 100 games of a superstar. 

I doubt that 10 days on the IL will change anything or they'd do it as you say. It's more likely take him as is or he's out until mid-August.

[Craig Arko]

Perhaps not.

[ashbury]

On 6/6/2022 at 5:36 PM, ashbury said:

He's had above league average OPS the past week and a half, so let's fix that by cherry picking away the good part?

Yo, sup y'all?     ??

 

 

/ disclaimer: any social literacy present in this post was provided by son Stashbury

[TouchEmAllGuy]

Time to retract the article? Lol 

Buxton seems to have found his groove again!!

[ashbury]

On 6/6/2022 at 7:35 PM, Nick Nelson said:

He's literally had one good offensive game in a month and I'm the one cherrypicking. Come on now. 

My comment to your article was a little flippant, but was intended to be constructive.  There was no further data at the time and I had said what I wanted to say, so I left it.  Today Buxton had an o-fer, so maybe it is an okay time to circle back to this example of data analysis.  By no means is my purpose to do an "I told you so" (as some seem to want) about Buxton's breakout right after this article came out, because I made no such prediction.  But I want to revisit the tangent about his week's statistics prior to the article.

No one, a week ago or now, disputes that Buxton had had a terrible several weeks following his injury.  An article looking back at the Twins' decision not to put him on the IL earlier in May could have been interesting, albeit second-guessing (which isn't always wrong to do). 

But the topic was whether to IL him "now" (a week ago).  The only thing a batting record can tell us about that question is a portion of his most recent games.  And that's hard because Small Sample Size is very risky to predict from, as we all understand.  You did look at this small sample, and to your credit you let it start when Buxton's string of o-fers ended.  But he'd had one good offensive game in a week - the month was no longer relevant if we were looking for evidence he was healing.  And his other games in that span were not a complete loss.  Good hitters can have a week like that even when they are 100% - as I pointed out, his OPS was even a tiny bit above league average for the period.

Sometimes data needs to be sifted, and outliers removed.  If Nick Gordon had been the topic of discussion, the game you wanted to discard from those recent ones might have made sense - that didn't look like a "Nick Gordon at his best" game, as it was too good - it would have been just a fluke in Gordon's season.  But this was Buxton, and that one game a couple Fridays ago was not beyond his ability when he's going good.  Tossing out that game was dangerous for a statistician to do without a very good rationale for doing so

Yours was a thought-provoking article.  But overlooking the one bit of contradictory evidence, that might suggest our favorite center fielder had found himself, was a mistake.  There's the old turn of phrase about "separating the wheat from the chaff."  You did that when you looked at Buxton's batting log.  And then, through bad luck or whatever, you threw away the statistical wheat, and continued to dwell on the chaff.

That was the point behind my quick comment.  I really think there is something to learn from this one.  And yes, if Buxton had gone an empty 3-for16 after your article, which could about equally well have happened, I probably wouldn't have thought to write this rejoinder now.   But I still should have.