Статистика Z-O подтверждает: 89% игроков возвращаются к исходному уровню после улучшения

The Normalization of Z-O: Can We Establish a New Baseline? Building on my piece a couple of weeks ago on the normalization of Z-O, it felt like there was a lot more left on the bone in terms of finding out if permanent changes to a hitter’s approach are possible, and if so, how long does it take for a new baseline to be established? In the first piece, we followed up on a discussion on X between redsoxstats and Eno Sarris, and introduced Z-O as a metric and the stickiness of it as a metric for plate discipline. It seems to be, in general, once a hitter establishes his baseline, it is difficult to deviate too much away from that level throughout his career. This is due to how inherent a hitter’s approach is at the plate. Hitters have tendencies baked into their profile, and shifting from that is difficult. But this opened up a slew of further questions relating to how baselines are established: is it possible to establish a new baseline? If a hitter improves or declines from this baseline, how are we able to establish a new Z-O baseline (if at all)? To test how much of a full-season Z-O change carries into the following year, I used full-season data from 2022 through 2025 pulled from the FanGraphs Statcast-ABS plate discipline leaderboard. I tracked each hitter’s Z-O across every qualifying season to establish their baseline mean or career average, then measured how far their Z-O deviated from it at any point during their career. A hitter who qualified in all four seasons gets a four-year average. A hitter who qualified in three, for example, gets a three-year average. The more seasons available, the more stable the baseline. To qualify for the pool, a hitter needed at least two prior qualifying seasons, enough to consider their level established. You could argue that for veterans like Freddie Freeman, who has been in the league since 2010, the baseline should stretch back further, but deviation from relatively recent data is more useful than tracking it back further, and almost four years of data is enough to capture a hitter’s established approach. That gives a pool of 205 hitters with an established baseline of at least 200 PA in a given season. For each hitter, I measured how far their 2026 Z-O moved from that baseline. A shift of +3 points or more from their career average classified a hitter as an improver, -3 or more as a decliner, and anything in between as flat. What was also problematic was when the biggest deviation from the average came in a debut season, or in 2022, the starting year, so I added an extra rule: if a hitter’s first qualifying season was the only reason they crossed the ±3 threshold in either direction, that flag gets removed. There is a small problem with taking the career average. This framework measures sudden shifts rather than slow evolution, so a hitter who has gradually improved his Z-O by a point or two each year will have that trend absorbed into his baseline rather than flagged as a deviation. That is a tradeoff worth acknowledging, but the questions driving this piece are about sudden shifts, not slow climbs, and the methodology reflects that. An absolutely crucial point that heavily affects the results of the Z-O analysis is that this year, in 2026, Z-O has regressed league-wide and pretty substantially. League-wide Z-O sat between 38.3 and 39.2 from 2022 through 2025, then dropped to 36.6 in 2026, the lowest in the dataset by over two full points. An article exploring potential reasons behind this is warranted, but worth flagging before we dive deep into normalization of Z-O baselines, as nods to potential decline may appear later in the piece. Of the 205 hitters, 69% were flat; their Z-O never deviated more than 3 points from their mean in any season. But what is even more interesting is what happened to those that improved or declined. 15% improved at some point but came back to their baseline by 2026. 7% declined and stayed down, and another 7% declined but recovered. Deviation in one direction has a lot more stickiness in its own right, as we will come to find out. This could be a long one, as I will categorize different groups of hitters and analyze the nature of their Z-O profile. The categories will go as follows: - The Non-Movers (69%) - Improvers Who Came Back (15%) - Improvers Who Held (2%) - Decliners That Never Came Back (7%) - Decliners Who Recovered (7%) Following this, I will try to answer broad questions about what we can learn from this and its application for our fantasy strategy. The Non-Movers The staying power, or stickiness, of a hitter’s Z-O is in full display here; at 69% of the pool, and almost five times the next category, with 141 of 205 hitters. Firstly, we should note, of these ‘Non-Movers’, how much movement actually was there within the 3-point scope. The mean deviation of the 141 hitters was 2.02 points of standard deviation (from their average. So generally in any given season, a hitter’s Z-O was 2.02 points away from their career average, which is not a lot. To put this in perspective, taking a rough league average, there are almost 4 pitches per PA on average, and 58% are thrown outside the zone, so roughly 1,040 outside pitches per season. O-Swing% is the share of those outside pitches the hitter swings at. 2.02% of 1,040 is about 21 pitches over a season, about one pitch every 8 games. Comparing how sticky Z-O is for a 2-year player vs a 4-year player, the range of deviation naturally widens with more seasons, but the standard deviation stays almost identical, from 2.21 points with 2-year hitters to 1.86 with four-year hitters. Additionally, the stickiness is not a function of how good or bad a hitter’s plate discipline is; the majority of this flat group sits between 30-40 Z-O, but it happens across the entire range, from the 3 hitters below 30 to the 3 above 50. There are 15 hitters with all four baseline seasons and a standard deviation of 1.15 or lower throughout their career. Xander Bogaerts (0.49), Nico Hoerner (0.64), Brandon Nimmo (0.77), José Ramírez (0.91), Christian Vázquez (0.92), George Springer (1.01), Dansby Swanson (1.04), Cody Bellinger (1.05), Spencer Torkelson (1.06), Juan Soto (1.07), Austin Riley (1.09), Ian Happ (1.11), Isaac Paredes (1.14), Bryce Harper (1.14), Freddie Freeman (1.15). But what is interesting is that in 2026, 7 of these 15 have actually regressed in their Z-O. Some of them pretty significantly: Christian Vázquez (-9.5), Austin Riley (-9.4), Bryce Harper (-5.5), Isaac Paredes (-3.8), Dansby Swanson (-3.3), Ian Happ (-3.3), Cody Bellinger (-3.0). These are established veterans who were locked in for four years, and even they could not escape the league-wide Z-O decline. Given how much this group has swung away from their career average, they will likely feature in the decliners pool next year. Emerging Decliners In this table, I have split a handful of players, the traditional non-movers, into three groups as a representation of the whole 141. 47% of the stickiest hitters in the pool are now declining beyond their established range. This is the middle group; even very established players like Ohtani and Riley are having career-low years. At the top, we have hitters that are incredibly sticky. The group moving up and above their career average, at the bottom, is much smaller. This trend of decliners heavily outnumbering the risers will become more and more apparent. Stickiness certainly has a bias in one direction. Currently as it stands, half of this pool of 141 hitters are emerging decliners. Pointing to the downward turn in this year’s Z-O across the league, it will be interesting to see where they move to next season. As we will come to find out, established, upper-echelon hitters are more likely to bounce back, whereas fringe or older bats are much more likely to continue to regress. Improvers Who Came Back A crucial factor when analyzing the deviation of a hitter’s Z-O is how sticky any deviation is; of course, for those that improve, we want to be able to see how this will hold. Well, the answer is, it doesn’t really. From this 2022 to 2025 range, 31 of 35 improvers came back to their career average; that is a whopping 89%. It is worth noting that this regression happens differently for everyone, with some players having consistent large fluctuations, where the stickiness is obviously less in effect, and some hitters with a less muddy reading. From this group, the stickiness of the hitters’ average Z-O really comes through; even those who have had a massive jump in their deviation from their mean Z-O, like Luis Garcia Jr., come back to their mean window the following season. In fact, it is much more likely that they will regress below their mean than they will sustain a new baseline. This also brings us back to the league-wide Z-O drop. Victor Caratini, for example, has a mean of 41.6 that spiked to 44.7 in 2025 and now has dropped to 33.4. A full season improver is likely to regress 89% of the time; this is how magnetic a hitter’s Z-O can be. And even with the league-wide drop this year, any improver that happens this season or any other season can be expected to fit right back in line the next year. Improvers Who Held The ones that broke through. This is a short list at just 2%, but is it proof it can be possible? Well, even that could be debated. Only 4 of 35 “improving” hitters sustained their new baseline, and none of them are particularly convincing. Rhys Hoskins has increased his metric from a career mean of 36.8 to 42.5 in 2026, but the improvement is relatively recent. After missing all of 2023 with an ACL injury, his trajectory was 37.5, 33.7, 40.2, and 42.5. The improvement began in 2025, giving us only one full season plus part of 2026 at the higher level. Lane Thomas has moved from a career mean of 36.3 to 42.0 in 2026. His Z-O sat at 34.7 and 33.8 in his first two seasons before jumping to 41.1 in 2024. He missed 2025 and returned at 42.0 in 2026. The higher level has therefore held for two seasons, although the gap year makes it harder to establish whether this is a sustained change. Henry Davis has increased from a mean of 37.9 to 42.6 in 2026. His only two baseline seasons were 33.8 and 41.6, with the metric increasing each qualifying year. But given the limited sample and his low wRC+, this may be a young hitter establishing a new baseline rather than evidence of a meaningful improvement, and one without fantasy relevance. JJ Bleday, the most interesting name of the 4, has increased from a career mean of 43.7 to 49.3 in 2026. However, his four previous seasons were highly volatile: 46.1, 48.9, 42.9, and 38.8, and even featured a year of regression. So I would take these with caution; for us to believe a player has established a new and improved baseline, even a full season isn’t enough, and some might argue never. As we will see, hitters can come back to their baseline and even regress very quickly. Even a full season’s worth of improvement is not sticky. Decliners That Never Came Back Of the 200-plus hitters, 29 of them actually regressed below their average at some point during the last 4 seasons. This decliner group can be split down the middle: decliners that were able to come back, reinforcing the stickiness of Z-O, and those that could not. These hitters showed regression during the baseline window, and that continued in 2026. For some tangible fantasy impact, if your hitter has a declining Z-O from his mean, alarm bells should be ringing. The average wRC+ of hitters from this group is 85. This pool mostly features hitters coming to the end of long careers like Altuve, Chapman, and Realmuto, or more fringe players like Heim and Gorman. Only Alex Bregman and Keibert Ruiz have a wRC+ above 100 from this group of 15. Worth noting, worst seasons from a Z-O perspective don’t necessarily correlate with a hitter’s least productive offensive season, but a persistent downward trend, more often than not, spells trouble. And while a full season improvement to a hitter’s Z-O is not sticky, a full season’s worth of regression is much more likely to hold. Decliners Who Recovered The other half of this declining group, and to me the most interesting group of names. These are those that declined but managed to recover back to their Z-O career average. This is a much stronger group of names than the one above, particularly from a fantasy perspective. It also gives hope to the emerging decliners that there’s a great chance they will revert to normal in 2027, especially if they are a productive hitter. For some hitters, a decline may be an intentional approach, as we have seen with Taylor Ward, for example. He sold out a lot for power in 2025, and the result was a career-high 36 home runs. This season he flipped the script and is focusing on getting on base. So his Z-O has come right back, and while he does not have the power numbers, he has the second most walks in baseball. Context matters, of course, but for the most part we can see the consistency of a baseline for each hitter across multiple seasons, even if there is volatility up or down. One of the most interesting names in this piece, I think, is Kyle Schwarber, given his profile as a high strikeout bat with power. He has been able to steadily improve on his plate approach, while staying within what would be considered his inherent tendencies as a hitter. Schwarber dipped to 34.0 in 2023 and has climbed every year since, sitting at 41.4 in 2026, nearly 3 points above his career mean of 38.7, with a 146 wRC+ to match. Mookie Betts is another example of not giving up on a hitter who may show regression even in the later stages of his career. Betts trended down for three straight years from 44.7 to 38.2 before snapping back to 42.6 in 2026. Summary The flat group shows Z-O is sticky when you’re at your level. Improving above it and staying there is a different question, and regression is the more likely scenario of the two to hold. I think the key takeaway when considering how sticky Z-O is as a metric is that 89% of improvers fell back vs the 52% of decliners that stayed. The most likely outcome is a hitter will hover around his average his entire career, but if the pendulum is to go one way, it’s towards regression. If a hitter does happen to improve even across a full season, deviating by several points away from his average, this cannot be taken as a new baseline. Schwarber, I would say, despite being in the last category, is about as good as you can expect from improving your Z-O. His average dropped in 2023, but since that dip he has climbed 7 points. Decliners, unfortunately, are much more likely to hold, and the strength of regression is much more stable than that of improvement. Z-O is indeed a sticky stat; the inherent tendencies of a hitter always seem to prevail. The league-wide downturn this year is something to keep in mind as we look to fantasy in future seasons. It will be fascinating to see if Z-O bounces back. Jack Martin is a contributor for RotoGraphs and also covers the Seattle Mariners for Last Word On Sports. Follow him on Twitter @jack_mariners.