Unlikely Pairs: Man Versus Fish

Most people in fantasy baseball have tiers of players. You likely believe that certain players are roughly interchangeable, and others are definitely superior or inferior to one another. Sometimes the differences between players appears so vast that it is obvious where you would draw the distinction, and other times it feels more fluid and dynamic. Intuitively you might feel there is a strong distinction between the two tiers, but it may be difficult to find a precise line in the sand.

In the past few days I have been going through my xStats, looking at where certain players fell on certain metrics. In doing so, I have noticed several pairs of players. Players who have performed very similarly on several different metrics. Some of these pairs, arguably, cross skill tiers. And boy do I love when players cross tiers.

So, maybe this will turn into a short series of articles. Let me know if you like the concept.

I will be quoting a number of stats which can only be found on xStats, so I’ll give you a brief explanation of these stats up front.

VH%, Value Hit Percent. The percent of batted balls which have an expected wOBA value of at least 0.88.  In 2017 only 6.1% of batted balls had a value this great. In the past, Value Hits have had a .879/.874/2.606 slash line. Yes, that is a 3.48 OPS, and a 1.424 wOBA if you’re curious.

PH%, Poorly Hit Percent. The percent of batted balls which have an expected wOBA value less than 0.088. In 2017 22.4% of batted balls fell into this category. In the past these balls have produced a .018/.018/.022 slash line. They are near automatic outs.

OUTs. This is a complete offensive metric, similar to wOBA, but using different input. wOBA is computed using singles, doubles, triples, walks, and hit by pitch. OUTs is computed using weakly hit balls, strongly hit balls, walks, hit by pitch, strikeouts, and expected home run rates. Lower is better, as the name implies, and the average score in 2017 is 0.106. The best batters in the league have a score of -0.100 or lower.

xAVG, xOBP, xSLG, and xOBA. These stats are identical to what you are used to seeing, except instead of using the actual hit totals they use the expected hit totals estimated by using the launch angle and exit velocity.

Alright, so with that out of the way, let’s look at the first pair:

Freddie Freeman v Mike Trout.

Yes, I’m starting here.

Both of these players missed about a quarter of the season with hand injuries. Mike Trout got injured on a slide, and Freddie Freeman was hit by pitch. In either case, hand injuries tend to limit power production after return, and these two are no exception. Indeed, these players were both putting up MVP calibre numbers prior to injury.

Trout hit a few more homers, but he also played an extra week or so prior to getting injured. Their production was almost identical.

It seems that Trout bounced back from his thumb injury slightly better than Freeman bounced back from his wrist injury. Which, maybe, should have been expected. It seems that wrist injuries might be a bit more serious than thumb injuries, but both of these guys put up tremendous numbers before and after injury.

That’s great, but you may wonder how much luck may be involved with these numbers. Especially with regards to Freeman, who lacks the track record of Trout. That is where the xStats comes in.

Power Production

With Trout, 31 xHR and 33 HR is a small enough difference that you can ignore it. But 28 to 35 is pretty large. Part of this may be an issue with park factors, since SunTrust is a brand new ballpark. Perhaps the home run adjustment is a bit too heavy handed? Maybe.

Here are a few of his non-home run batted balls ranked by xHR%.

98.8% chance for a homer! Well, obviously I have to include a gif. Be warned, it didn’t involve some miracle catch or anything flashy.

Well, this looks like a routine flyball, how could that possibly have a 98.8% home run rate? Some details:

This hit occurred on a cool-ish (67 degrees) April day. It was an opposite field hit, so it may have been on the low end of the backspin spectrum. It is possible that the wind knocked it down a bit. However, the ball was hit 101 mph with a 31 degree launch angle (and -35 degree horizontal spray). There have been 16 balls hit in this manner, and 15 of them were home runs (94%).

If I expand the search a little bit (+/- 2 on each of these three parameters) I find there are 95 batted balls, 88 of which were homers (93%).

These chances are boosted to 98.8% by a combination of factors including but not limited to: park factors, decreased average coefficient of drag in 2017 compared to 2015, and a result of the ‘most similarly hit balls’ algorithm I have in xStats. However, with or without those adjustments this batted ball is almost always a homer, even if it looks like a mundane fly ball in this gif.

Getting back to the high home run probability table, you’ll notice 7 of the 10 listed were hit during away games. So, perhaps the SunTrust park factors aren’t a huge issue here. This list of 10 is only a small fraction of the overall home run probability table, and it does extend further with many more balls in the 40-50% range.

The xSlash

These players produce offense in a slightly different manner. Freeman appears to have higher slugging, while Trout has higher OBP. We’re splitting hairs here, though, as both players are tremendous batters in every respect.

The expected slash numbers are an attempt to wipe away luck and examine the underlying skills. The high probability home run above is a great example. That ball almost always leaves the park, but Freeman’s got caught. That’s too bad for him, but at the end of the day he produced a batted ball with a ridiculously high probability for success, and he should be rewarded for that.

Traditional stats give him a zero, xStats gives him .988 of a home run. Of course this is a double edged sword, and a xStats can subtract low probability hits as well If you don’t believe me, go look at Didi Gregorius’ xHR totals.

These slash line items revolve around the expected singles, doubles, triples, and home run rates. You can take a step back and generalize batted balls even more by stating “hard hit” or “soft hit”. You can see that sort of generalization in the table below.

Earlier on I described how Value Hits are near automatic extra base hits, and Poor Hits are near automatic outs. In 2017, Freeman had more Value Hits and fewer Poor Hits when compared to Trout. However, Trout had fewer strikeouts and more walks.

In other words, in terms of “hard hit” and “soft hit”, Freeman comes out on top. He has more hard hits and fewer soft hits. However, walks and strikeouts are obviously hugely important.

As I mentioned before, OUTs combines the strong and weak hits with strikeouts, walks, and expected home run rates. The strong and weak hits are not defined by Value Hits and Poor Hits, but they are similar. You can read about it here. Anywho, when these are all combined, the OUTs stats for both players is almost identical, suggesting that they provide almost the same amount of offense to their respective teams when all is said and done (disregarding base running).

In fact, knowing the bbFIP constant for 2017, 5.3958, allows you to convert the OUTs stat to a “Runs per game” metric. You only need to multiply OUTs by -11 and add the constant.

Freeman: 7.64 Runs per Game (bbFIP)

Trout: 7.58 Runs per Game (bbFIP)

That means over the course of a 162 game season, a team of Freeman’s would outscore a team of Trouts by about 9 runs, if you disregard base running. If you factor in Trout’s extra speed and base stealing threat, the scales would tip into Trout’s favor.

Going Forward

Mike Trout has been the best player in MLB since his debut in the majors. He is, already, an Inner Circle Hall of Fame player. There is no doubt about that. Mike Trout’s floor as a player, assuming he is healthy enough to play even a solid fraction of a season, is still better than just about everyone in baseball.

I am not trying to argue that Freeman is a superior overall player. However, I am trying to argue that Freeman is, at least, close to Trout in terms of value as of this moment. Trout was better than Freeman 3 years ago, and he might be better than Freeman 3 years from now. But let’s live in the moment. Just today. How much better is Trout over Freeman, realistically? In 2017, for example.

They finished with roughly the same batting average. Trout had 5 more home runs, and 14 more steals, 8 more runs, and 1 more RBI. We can ignore the injury problems, since they both had similar injuries missing similar time. So, yes, Trout was better, especially in terms of steals. But Freeman was within striking distance, and Freeman appears to have had a somewhat unlucky season.

The mock drafts to date place Trout as the undisputed #1 overall pick, where he belongs. However, Freeman is ranked 20th overall.

I have a projection for these players, it is based upon all three years of data I have for these players.

Freeman’s higher average and slugging, and Trout’s higher OBP persists through this projection. However, their overall offensive contribution, measured in both wOBA and OUTs, is nearly identical.  Trout will likely steal more bases. Those steals will be worth quite a bit.

Of course, Freeman isn’t the only great first basemen. Paul Goldschmidt is similarly great, and even superior to Freeman in a few ways. Namely, a slightly higher OUTs projection going into 2018, and of course the superior base stealing threat. There is a great caveat with him, though:  The Humidor. If a humidor is installed in Chase Field, and rumors suggest it could be on the horizon, then Goldschmidt’s power numbers immediately come into question. He could lose a large number of home runs, perhaps 10 or more.

In my mind, Freeman is a first round talent. I personally slot him into the 14 ADP slot between Carlos Correa and Chris Sale. If the humidor is installed in Chase, I would drop Goldschmidt out of the first round and push everyone up a slot.

What’s your take? Do you like the idea of comparing pairs of players? I have a few other names who could receive a similar treatment.