Twins Tidbit: Offensive Stock Report

[TwinsData]

As we move in August, which Twins hitters are heating up and solidifying their place in the every day lineup? Which are cooling off and casting doubt as to their role going forward? Let's take a look at the trends to determine which stocks to buy and which to sell.

 

The above graph shows rolling OPS (on-base plus slugging percentage) over the past 30 days for each hitter with a sufficient number of plate appearances in the last month.  For example, the height of the line at 7/1 indicates the hitter's OPS in the month of June. Hitters are sorted by their current OPS in the last month from left to right and then top to bottom. The dashed line indicates the MLB-average OPS of 0.708.

Some takeaways:

• Jose Miranda is the hottest hitter in the Twins lineup with a 1.054 OPS in the last 30 days, capping off a meteoric rise. On May 15, his rolling OPS was a paltry 0.322. Stock way up.
• Jorge Polanco is similarly on fire. His rolling OPS has been above 0.900 for almost the entire month of July.
• Ryan Jeffers' injury is a shame. He had really started to heat up. At the same time, Gary Sánchez has been in a steady decline since June began.
• The consistent performance of Luis Arraez is remarkable. He has yet to have a 30-day stretch with a below-average OPS. By contrast, Byron Buxton is the most volatile hitter in the lineup with 30-day stretches of both 1.360 and 0.532 OPS.
• An underreported part of the Twins' forgettable July has been the cooling of Buxton (0.643 OPS in the last 30 days) and Carlos Correa (0.702). Miranda and Polanco were a band-aid over a struggling top of the lineup.
• It was past time for Gilberto Celestino to return to Triple-A. He owns a 0.488 OPS in the last 30 days and had been trending in that direction for some time.

What else can we take away from the hitting trends that might help the Twins construct a winning lineup down the stretch?

 

View full article

[Karbo]

Miranda looks like the real deal. It looks to me like we have some pretty good youngsters proving they can hit. Another is Larnach, hopefully he keeps hitting when he comes back. So 3 rookies making themselves valuable major leaguers. Now we need pitching, and lots of it!

[ashbury]

Love this form of data visualization!

[wabene]

14 minutes ago, ashbury said:

Love this form of data visualization!

Who is this mysterious Mr Data? 

[PatPfund]

From someone who went all rant on your last piece, I really liked this. Good data, and effectively presented (reinforcing some general impressions I had, but the Jeffers graph in particular caught my attention. Longer upswing than I was giving him credit for.

[wabene]

This is a really great idea. I'm sure the Twins analytics department has gobbled it up already like the Twins Daily Bullpen Usage Chart ©

[stringer bell]

Well done. Much appreciated.

[Mike Sixel]

Good stuff, thanks.

[ashbury]

Okay, quick question, what does the horizontal dotted line represent?  I assumed it was league average OPS.  But Alex Kirilloff's OPS for 2022 currently stands at .669 which translates to a below average OPS+ of 92, while in his graph above, he is above the dotted line for every segment of this season.  It also looks like his clock started in late June, in that graph, but his overall season numbers have been dragged down by his abbreviated April/May.

[TwinsData]

1 hour ago, ashbury said:

Okay, quick question, what does the horizontal dotted line represent?  I assumed it was league average OPS.  But Alex Kirilloff's OPS for 2022 currently stands at .669 which translates to a below average OPS+ of 92, while in his graph above, he is above the dotted line for every segment of this season.  It also looks like his clock started in late June, in that graph, but his overall season numbers have been dragged down by his abbreviated April/May.

Correct. The dashed line is league average OPS.

Those periods are removed from the data for Kirilloff. He had 18 plate appearances in April, then was out for 3 weeks, then had 14 plate appearances in May, then was out for 5 weeks. To include periods like that in the graphs would be distracting and misleading both for sample size reasons and the fact that there is no continuous stretch of play to actually measure a trend.

[tony&rodney]

What would help the offense be productive through the remainder of the schedule is to add some significant pitching and release the pressure of scoring 7-10+ runs every game. That said, the lineup just needs to eliminate all distractions and go pitch by pitch and wear down their opponents.

[Richie the Rally Goat]

Great work TwinsData!

graphical evidence Max Kepler is “Meh” and Miranda is a monster! Imagine if more of the hitters could coordinate the upswings… the runs to be scored….

[ashbury]

On 7/29/2022 at 4:18 PM, TwinsData said:

Correct. The dashed line is league average OPS.

Those periods are removed from the data for Kirilloff. He had 18 plate appearances in April, then was out for 3 weeks, then had 14 plate appearances in May, then was out for 5 weeks. To include periods like that in the graphs would be distracting and misleading both for sample size reasons and the fact that there is no continuous stretch of play to actually measure a trend.

No disagreement really.  I expect that there are several in the TD readership who have had reason to think carefully about aspects of data visualization during their education or careers.  Some cleaning of data is often called for.

I hope you'll still give some thought to the underlying issue I raised, though.  Kirilloff's season numbers show a slightly below average OPS, but his chart fails to give any indication of why that might be.  Even though the omitted data is a small sample, it was so poor that it accounts for the difference.  Maybe having the lower cutoff of the y-axis be .000 instead of something around .300 is all that's needed.  A thick line to represent sufficient amount of data for a trend line, a thinner line to represent less sufficiency.  Those kinds of tactics.

[Minny505]

Hands down my favorite regular read (at least I hope it will become that) on Twins Daily, which is a huge compliment.

Being a new way to look at data, I also want to make a suggestion. 

Changing the league average OPS to a rolling average as well. The OPS being a flat bar, particularly in a season like 2022, can be deceiving. Having a rolling league average would show us individual trends vs league wide trends.

Or put them both on the graph. Regardless, I'll be digging in every time you publish.