Stats Must be put in context

Earlier on this week I saw a tweet that annoyed me; I know shock of horrors that a user gets annoyed on Twitter. It wasn’t the baseball usual of someone defending a wife beating player/owner or saying minor league players should be happy with what they are paid because they get to play the great game of baseball.

It was someone using some stats to say that a player was bad, and they wanted them to perform better. The offending tweet is below. I don’t follow this individual so therefore some I do liked or retweeted this for it to hit my sensitive eyes.

The average fan would look at those numbers and think Stanton performed poorly, in 2018, in two strike counts. But since there is no comparison to another player or league average, how is someone reading this supposed to infer anything else? That’s what annoyed me.

It takes a bit of time to look up a player’s stats though each count of an at bat. So if this user wanted they could have provided context. They didn’t. So for me the purpose of this tweet is to make you think Giancarlo Stanton was bad in 2 strike counts and they using stats to try and back it up.

There is a famous book in the world of stats called ‘Lies, Damn Lies, and Statistics: The Manipulation of Public Opinion in America’. It was published in 1976 and it is as relevant today as it was back then. Context is king. When providing stats if you don’t provide any context your either taking an assumption that the context is so obvious it need not be added or you are trying not to add the context as it will moot your point.

So in my mild mannered British rage, sent a reply showing what the league averages were in these 2 strike counts and the wOBA for them as well. As this stat is a much better measurement of a batter’s performance.

Unsurprisingly it showed that Stanton, when we use wOBA, is much better the average for the MLB in 2018. Suffice to say I did not get a response from the original tweeter but I hoped that anyone who would click on the tweet going forward might see the context that is required.

With context provided I felt my job was done, I had saved the day and could continue with my tweeting about my fantasy baseball drafts, but a day later the tweet was still annoying me.

Of course Stanton is better than average, I would have been shocked if he wasn’t. The real question is he how well does he perform compared to his peers, the likes of Aaron Judge, Mookie Betts, Mike Trout, in these scenarios. Are these numbers good for one of the best hitters in MLB?

Thankfully I have done some analysis so you don’t have to and I can answer that question. Let’s start with the MLB averages.

The table above shows the outcomes stats for all at bats in 2018. It is split by count, i.e. any at bat which had an pitch in a 0-1 count would be in that row, even if the at bat didn’t end on that pitch. 0-0 is for all at bats and you can see the average for all at bats regardless of count. Some simple things pop out of this chart, the outcome is on average worse, compared to the 0-0, when a player is level or behind in the count and better when they are ahead in the count.

This is our base performance level. Let’s look at Stanton. I have included a rank, for each count, for his wRC+ against the 448 hitters last season who had 100+ plate appearances so we can compare him to other hitters.

Giancarlo Stanton 2018 MLB Stats

There is a lot to unpack here but I want to focus on a few key points.

  1. Stanton’s overall wRC+ ranking was 51st, at 127, this is already lower that most Yankee fans were wanting for 2018. He didn’t have the same level of performance as he did the year before.
  2. He is better than league average in all counts.
  3. His wRC+ for 0-1 and 1-0 are similar to his overall wRC+, he doesn’t show the same percentage decrease or increase as the overall population. It is as if the first pitch almost doesn’t matter to the outcome of his at bats.
  4. After the 0-1 count, his next 3 worst rankings are all 2 strike counts.
  5. He gets into a higher % of 3 ball counts than average. Where the wRC+ is highest, which in turn boost his overall ranking.

This might lead you to say that Stanton performs worse compared to other hitters when in a 2 strike count. I think it is fair to say that but I feel we need to compare this to other power hitters to be certain. He isn’t going to have the same profile as a Joey Votto or Jose Altuve. Let us look at J.D. Martinez and Aaron Judge to see how Stanton compares.

Both players ranked higher than Stanton overall, but when looking at the 0-1 and 0-2 counts Stanton performed better than both of them. Judge stands out as being strong in 1-1 count with his performance better through here than through 0-0 and Martinez is very strong when he is ahead in the count.

Both of these power hitters see significant drops in their respective rankings when in pitchers counts but it is their performance when ahead or level in the count which separated them from Stanton in 2018.

The context which this adds by no means gives us as direct answer but I believe Stanton was simply not as good as these guys last season and I don’t think it was down just his performance in 2 strike counts. He didn’t punish pitchers enough. when he got ahead in the count.

Going forward, I hope you all will now question a stat if your aren’t give context.

That all being said. Now as I have spent the time collating this data, we might as well have a look at some of the other top players just for fun.

I have highlighted these 6 because they all show something very interesting.

  • Mookie Betts may have been too aggressive in batter counts, his batting average for at bats through 3-0 was significantly worse than others. Small sample size caveat.
  • You had to pitch a strike to Shohei Ohtani on the first pitch, when behind 0-1 he was worse than league average. But when 1-0 he was second only to Mike Trout.
  • Jose Ramirez was the most consistent across all pitch counts, his wRC+ ranks deviated the least out of all the 100+ AB hitters.
  • Juan Soto was very good when down in the count, this comes from the phenomenal plate discipline that he showed.
  • Mike Trout and Mookie Betts both performed better than average at bat regardless of count. So Trout and Betts down 0-2 are better than the average hitter at 0-0.
  • Joey Votto, shows low wRC+ ranks in most counts but his higher percentage of counts with 3 balls and that is why is overall ranking is so high.

If you want any other charts hit me up on twitter @Reassom and I endeavor to reply. (100 PA min)

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.