clock menu more-arrow no yes

Filed under:

Looking at the Jays Hit Probability Numbers Through Half of the Season

New, comments

Some good news, we might be due for some positive regression...but we also might not be, who knows?

MLB: Toronto Blue Jays at Kansas City Royals Peter G. Aiken-USA TODAY Sports

Half of the MLB season has gone by for the Blue Jays. Considering the poor results, it was a half we wouldn’t like to look back on too much - so let’s do exactly that.

For those that didn’t catch my initial article at the quarter point of the season, I built a model to gauge hitter’s ‘luckiness’ based on Statcast’s new publicly available ‘Hit Probability’ metric that has been released this year. I recommend reading it here, as it explains in detail what hit probability is, details of the model, the limitations, and how the numbers were generated.

Simply though, each hit or out in play is assigned a percentage of it’s likelihood of being a hit based on its exit velocity and launch angle of the ball. Essentially, if a player gets a hit that is graded to be a hit 70% of a time, the model will credit them with a 0.7 hits, and if they get an out that is graded to be a hit 20% of the time, the model will credit them with 0.2 hits. It uses these numbers to calculate what I call their adjusted batting average, and I have added walks to these numbers to create their adjusted OBP. I will be analyzing the gaps between their actual OBP and their adjusted OBP to see who has been getting lucky or unlucky.

A Quick Note on Running Speed

As addressed in the initial article, I highlighted baserunning speed as a something that may affect the results of this model, as faster players can convert infield hits more often than slower ones. With recent reports from Statcast highlighting the Jays as being an especially slow team (if not the slowest), this is almost certainly part of their ‘unluckiness’ when it comes to beating their adjusted OBP.

Statcast’s description of the new foot speed metric: “feet per second in a player’s fastest one-second window.” The Major League average on a “max effort” play is 27 ft/sec, and the max effort range is roughly from 23 ft/sec (poor) to 30 ft/sec (elite). A player must have at least 10 max effort runs to qualify for this leaderboard.

The First Quarter Results

The First Half Results

The Unlucky

Kevin Pillar - Look here for some much needed Kevin Pillar optimism. His adjusted OBP has fallen by .009 which shows Kevin has been hitting the ball almost as well as he was in the first quarter, he just hasn’t been getting the results from it. Kevin’s above average speed (27.8 ft/s) marks him as a guy who could even consistently beat his adjusted OBP, so Kevin could be due for some positive regression from his .034 under-performance of his adjusted OBP right now. He’s been largely the same hitter, just getting much worse results out of it.

Devon Travis - Before going down with a potentially season ending injury, Travis continued his trend of closing the gap between his actual and adjusted production. His adjusted OBP remained almost identical, while his actual OBP rose by .040. Travis’ jump in production has been a prime example of what this model can predict, and while he didn’t have enough time to completely close the gap before going down with an injury, he was well on his way to doing it.

Russell Martin - Martin has been getting much better quality contact since the quarter-season mark, improving his adjusted OBP to .393 from .371. Unfortunately, his actual OBP has only increased by .004. With a pessimistic look, his below average speed (26.2 ft/s) could signal this to be the level of contact he needs to truly sustain this OBP in the future, but on the other hand he could be in line for even more positive regression, given his results from the first-quarter.

Luke Maile - The positive regression has already started! While his OBP has moved up to .149, he is still well off of his adjusted OBP of .211. His adjusted OBP is actually down from the .248 it was at during the quarter mark. He’s actually been hitting the ball worse. He could still be in line for some positive regression as he is .062 below his adjusted OBP.

Barney, Goins, & Coghlan - Could see some slight positive regression, they range from .017 to .026 under their adjusted OBP.

The Near Where They Should Be

Justin Smoak - Some of you may have thought I was crazy when I placed Justin Smoak in the unlucky category in the quarter-season update and said he might be due for even more positive regression. Well, that positive regression came. Since the quarter-mark, Smoak’s adjusted OBP has stayed around the same, while his actual OBP has increased from .342 to .370. He was making outstanding contact and getting good results, now he’s getting outstanding contact and getting outstanding results. He’s still under-performing his adjusted OBP by .026, and that’s likely due to his team lowest 24.7 ft/s speed, so Justin has been getting pretty much the results you’d expect from the contact he’s been making. Justin Smoak has not been a fluke.

Josh Donaldson - While an injury sidelined him in the first quarter of the year, during his small sample the deemed his high .429 OBP to be unsustainable. Unfortunately, the model was right. Since then in a much larger sample, Donaldson has got on base at a rate of .355, which is only .002 away from what the model had adjusted for him at the quarter mark of the season. The .013 difference now could just be a product of his below average speed (26.0 ft/s).

Jose Bautista - Jose has seen positive regression since the quarter mark. His OBP is up (.342 from .330), and is approaching where the model expects him to be at (.353). The difference can likely be explained by his foot speed at this point (25.3 ft/s), so he’s probably gotten the results that have merited his performance so far.

Kendrys Morales - Kendrys has seen positive regression since the quarter mark. His OBP is up (.309 from .286), while his adjusted OBP has fallen only .009. Now, his actual OBP is still .036 lower than his adjusted average, but I think it’s time to put heavier emphasis on footspeed in this model. It’s no secret that Kendrys is one of the slowest players in the league (24.7 ft/s), and that should be enough to say that Kendrys’ under-performing his OBP isn’t a result of bad luck, but rather bad foot-speed. Kendrys is getting on base at the rate he would be expected to based on his contact, given his speed.

Troy Tulowitzki - Unfortunately, I’m not here to bring Tulowitzki optimism. Tulo hasn’t seen many jumps in his actual or adjusted OBP since the quarter mark despite tripling his sample size. His OBP still sits .028 away from what is expected from him, but this is probably boils down to his well below-average speed (25.4 ft/s).

Steve Pearce - Since the quarter mark, Pearce has both massively increased his production (.256 to .323 OBP) while also improving his hit quality and increasing his adjusted production (.295 to .327). Not only is he getting the results he deserved at the beginning of the year, but has also exceeded them and got the results to match. He sits only .004 away from his adjusted OBP at this point.

The Lucky

Ezequiel Carrera - Zeke continues to beat his adjusted OBP. Due to the speed developments, I was tempted to put him in the right where he should be category. Another quarter of beating his adjusted OBP and I just might. Zeke has seen a spike in his OBP with a large increase in walks this quarter, but saw his batting average by .020, likely due to negative regression. He now sits .030 higher than his adjusted OBP, which could very well be due to his well above average speed (28.1 ft/s, fastest on the team).

Dwight Smith Jr. - It’s not a shocking revelation that a guy with a .588 BABIP may have been benefiting from some good luck. The question is, how much of that could actually be sustainable. With a small sample of 16 batted ball events though, the model shows he was actually making pretty good contact. While his actual OBP was .046 over his adjusted OBP, this shows that he might be in line for some negative regression. With his positive speed though, there’s a chance that he can consistently outperform his adjusted OBP, albeit likely by a smaller amount. We’ll need a much larger sample size to say for sure.


Some of the predicted positive regression has come true. They’ve got on base at a rate .011 higher than the first quarter. They have also increased their adjusted OBP to .336, so they are making better efforts at getting on base and have partially been rewarded for it as well. Unfortunately, they have probably gotten the results they should expect given their slow speed. They’ll need to make some fundamental improvements in the second half if they want this offense to improve rather than hoping their bad luck runs out.