Introducing LASR, Providing Much-Needed Context to Your Favorite Baseball Statistics
Please visit this Tableau story to see the project referenced in this article.
Hello! My name is Tucker Atwood and this is my first plate appearance on High and Outside, a Substack newsletter that I am debuting right now. I recently switched career paths and got my Master’s in Data Analytics, and one way I want to use that degree is to develop a baseball analysis portfolio. So, that’s what this is. I am also a playwright, author, actor, former math teacher, have thru-hiked the Appalachian Trail, and bike-toured the East Coast Greenway. More information about me can be found here.
This first High and Outside post is about a project that contextualizes baseball stats to make them easier to understand. I’ve written a short and long explanation of the project below. The project itself can be found on the Tableau story linked above. Enjoy!
Short explanation:
League-Adjusted Standardized Rating (LASR) stats are designed to provide context for player production, to more efficiently identify player quality and player type. To do this, all ratings are placed on the same scale, which is the traditional scale that scouts use: 20-80 with increments of 5. For those unfamiliar, “50” represents average, and every 10 points above or below 50 is equal to 1 standard deviation. For example, a hitter whose OBP rating is 75 gets on base at a rate 2.5 standard deviations above the mean. These means and standard deviations are established by qualified players from that particular season.
This allows us to better understand a player’s performance in comparison to his peers. With MLB teams constantly aiming to develop the next big strategic advancement, the league scoring environment shifts at least slightly every year in a reflection of new approaches and players. With these changes comes more emphasis on historically ignored statistics like BB% and ISO, which even diehard baseball fans often lack a preconceived notion of quality for. Without a fundamental understanding of the context in which these numbers appear in their endless scroll across our screens, we have no choice but to ignore them completely or use them recklessly in our calls to overthrow front offices.
There are already some established stats, like OPS+ and ERA+, that aim to do the job I’m doing here. However, they do not take standard deviations into account and therefore the scales for those adjusted stats are not consistent with each other. There are also certain adjustments and details within the LASR calculations that are important to note, but I’ll save those notes for the longer description. If you’re skipping that, my advice is to just go play around with the graphs and figure things out for yourself. Have fun!
Long explanation:
“This guy’s batting .237/.321/.457 this year with a dozen homers, 58 ribbies, and 60 runs. He strikes out at a 16.4% clip and walks 9.2% of the time. He’s averaging a career-high 0.16 sacrifice flies per game, his current heartrate is a cool 137, he slept 7.23 hours last night, he wet his pants 324 times as a toddler, and he got an 84 on his third grade science project. He takes a ball, high and outside, 1-0 count.”
My love of baseball has always revolved around the stat sheet. When I was younger, I collected Bill James Handbooks like my friends collected girls’ phone numbers. I would pore over the endless pages of stats as if they held the secrets to the universe. My favorite activity was scanning players’ statistics and envisioning the ebbs and flows of their careers: young superstars who fizzle too early, late bloomers who defy Father Time, bat-to-ball contact hitters picking up some power, wild flamethrowers finally discovering the strike zone, one-year deals booming, mega contracts busting, injury-riddled seasons, inspiring comebacks – and on and on. Though my statistics dealers have shifted to the likes of Baseball Reference, FanGraphs, and Baseball Savant, I still do this. Why shouldn’t I? More and more baseball happens every year! And as the numbers pile up, so do the stories hidden between every digit, waiting to be discovered.
The problem is, these beautiful stories are indecipherable without context. Though I’ve followed baseball my whole life, when a string of statistics is presented on wildly different scales with no reference points and several qualifiers, I have to make a serious effort to either: take a desperate swing trying to make contact and understand the data, or blink and let it go by for a strike. This is true for both traditional and newer sabermetric statistics. Sure I know all about batting average, but is .256 good these days? Yes I know a pitcher’s walk rate could explain or predict success, but is this guy’s 7.4% walk rate good, bad, or boring? Maybe I’m familiar with expected and weighted stats, but are you really going to just shove a .361 xwOBA down my throat without letting me chew on it a bit first?
This gap between the abundance of and comprehension of presented statistics is an issue not only for lifelong fans like myself, but also for new fans, for whom an endless stream of pointless values serves only as an unnecessary barrier to entry. If I, a seasoned baseball nerd, am struggling to comprehend the significance of some batted ball data, then casual fans must be either watching it fly over the wall or throwing it into the stands themselves. I don’t want newcomers to give up on baseball before the game itself can pull them in, and I want existing fans to better understand the stats they so often misrepresent in bar conversations and online arguments. For everyone’s sake, we need some context.
LASR is here to help.
Ultimately, a player’s success is measured in relation to their peers. In 1924, Dazzy Vance struck out 21.5% of the batters he faced, a rate which Mitch Keller posted in 2024 when he ranked 37th out of 58 qualified starting pitchers. If Dazzy put up his numbers today, he wouldn’t be anything special unless he made up for it elsewhere in his game. Luckily for him he posted those stats a hundred years ago, when he easily led the league in strikeout rate and made the average pitcher look like a tee-ball stand. Of course, if you are familiar with trends throughout MLB history, you can remind yourself that strikeout rates have drastically changed and understand that Dazzy Vance and Mitch Keller are not the same type of pitcher. But there’s no harm in making player evaluations a little easier, right?
League-Adjusted Standardized Rating (LASR) statistics put all the baseball stats we know and love into the proper context. Here is how they are calculated:
For each individual season, I have used all qualifying players to establish the means and standard deviations of selected rate statistics that I have deemed important in understanding a player’s value and type. Qualifying thresholds are commonly used in MLB for rate stat leaderboards because otherwise small samples would wield a much heavier sword than they should (“Wow, this guy went 1-for-1 in his debut and now he leads the league in batting average!”). Similar quirks would happen with LASR stats if I included all players in the mean and standard deviation calculations, so using qualifiers to establish baselines was necessary.
For hitters, qualification equates to 3.1 Plate Appearances per Team Game, which comes out to about 502 PAs in a typical 162-game season. For starting pitchers, qualification is 1 Inning Pitched per Team Game (162 IP in a normal season). Though I couldn’t find as established a qualifying threshold for relievers, a general rule of thumb seems to be 1 Out (0.1 IP) per Team Game, or 54 IP over 162 games, so that is what I have used.
For all players in that season, I have measured their z-score in that statistic. A z-score is the number of standard deviations away an individual value is from the mean of the set. For example, if the mean batting average of all qualified hitters for a season is .273 and the standard deviation is .027, then a player hitting .300 would be 1 standard deviation above the mean, while a player hitting .327 would be 2 standard deviations above, and so on.
For some statistics, it is preferable to have a lower value; at this point, those z-scores are multiplied by -1 to reverse their direction so that all “better than average” values are expressed in positive z-scores, and vice versa. This includes most pitching statistics, such as ERA, FIP, and WHIP. For some outcome stats like strikeout rate and walk rate, I have renamed them something like “Avoid K%” (for hitters) to show that the skill, avoiding strikeouts, is being measured with positive results in mind – i.e., a hitter with a low strikeout rate will have a high Avoid K% rate.
For non-qualifying players, I have adjusted their z-scores by a scale based on their progress toward the qualifying threshold. For example, if a player is hitting .327 (a z-score of 2, remember) but only has half of the plate appearances they need to qualify, their adjusted z-score is 2 * ½ = 1.
These adjusted z-scores are translated to the 20-80 scouting scale described above in the short explanation. This is done by multiplying the z-score by 10 and adding the result to 50. So, our .327 half-qualified hitter now has a batting average LASR of 60.
Scouting reports conventionally are presented in increments of 5, so these adjusted z-scores are then rounded to the nearest 5. This strips unnecessary noise from the data so we don’t get too bogged down on the difference between, say, a 56 and 57 – essentially, differences these small are likely due to luck and are not a genuine indication of one player’s true talent over another. This simplifies the process of comprehending LASR and allows us to sort players into easily understood “categories.”
That’s the LASR system in a nutshell. To summarize:
Calculate the mean and standard deviation for all qualifiers that season.
For each individual value of all players that season, subtract the qualifier mean and divide by the qualifier standard deviation.
For non-qualifying players, adjust the value by a scalar equal to their progress to qualification.
Multiply the adjusted value by 10 and add 50.
Round to the nearest 5.
And if you prefer a formula:
Let’s use the rookie phenom Paul Skenes as an example. He spent the first month or so in the minors and made his MLB debut on May 11th, then pretty much immediately cemented himself as one of the best pitchers in baseball. Because of the late start, he only pitched 133 innings in 2024, which is below the qualifying threshold, but in those innings he posted a 1.96 ERA. Of course most baseball fans, and even casual observers, probably know that’s good, but how good is it? And how can we measure this production while taking into account his somewhat low innings total?
Among qualifiers, the mean ERA in 2024 was 3.7099, with a standard deviation of 0.6838. So, the ERA Skenes posted (1.9624 to keep decimal places consistent) had a z-score of:
For ERA it is better to have a lower value, so this z-score is then multiplied by -1 to give us 2.5556. This means Skenes was about 2 ½ standard deviations above average. However, he didn’t have enough innings (133) to qualify (162), so the next step is to multiply by his progress toward qualification:
So, he gets to keep most of his impressive season, but he is fairly knocked down for not quite getting to the threshold. Now all we have to do is express this as a scout would:
Which, rounded to the nearest 5, becomes an ERA LASR of 70.
And that’s how all LASR values are calculated! This puts all stats on one consistent scale that’s easy to understand. As mentioned in the shorter description, there are already league-adjusted stats like OPS+ and ERA+ that do what I’m aiming for here; however, standard deviations are not accounted for in those calculations and all values are presented as a percentage better or worse than league average. While that may be helpful, I believe accounting for the spread of data is necessary in comprehending league-adjusted statistics and comparing different statistics to each other. I’m not trying to pick a fight with the sabermetricians; just clarifying why LASR provides something that doesn’t already exist.
Plus, I can apply LASR to all stats that are publicly available, not just a select few. Since this is the first iteration of this project and I didn’t want to clutter the dashboards too much, I have limited the scope of LASR for now. But don’t worry: I have the capability of including more advanced statistics, and plan to in the future. For now, here are the stats I have selected to include:
Hitting statistics:
OPS
AVG
OBP
SLG
HR%
BB%
Avoid K%
ISO
BABIP
DRS
wBsR
Pitching Statistics:
ERA
FIP
WHIP
K%
Avoid BB%
Avoid HR%
Avoid BABIP
GB%
Called Strike%
Swinging Strike%
Zone%
If you are unfamiliar with any of these and/or why I’ve included them, I encourage you to research them a bit – or reach out to me and I’ll give an explanation. Generally, I want to provide an overview of a player’s quality and type. For example, Juan Soto and Shohei Ohtani are both excellent hitters, but they differ in the way they provide value. Just looking at their raw stats might provide some insight into their differences, but putting everything from BB% (Soto: 85; Ohtani: 60) to wBsR (Soto: 40; Ohtani: 95) on the same scale helps tremendously in comparisons.
I’ll wrap this up with some housekeeping notes:
Every part of this is a work in progress. I am releasing it into the wild right now because I think it’s useful in its current form, but that doesn’t mean I think it’s perfect. I hope to carve out enough time in my schedule to continue working on this and making it as helpful as it can be. This means I’m open to suggestions from everyone, from baseball fans who will clamor for xwOBACON to be added, to my friends and family who have been lost throughout this entire explanation.
You may notice that I had a tough time finding a workaround for players with identical names, but I did find a solution. To avoid mixups, I added a (2), (3), and so on to players who debuted with the same first and last name as a previous player. This also applies to players with suffixes like “Jr.” because those didn’t show up on the datasets I used. So, yes, Bobby Witt Jr. now goes by Bobby Witt (2). Maybe I will find a better workaround, but this works for now.
Tableau is a great data visualization tool, but I’m not sure if it is the best home long-term for this project. There are certainly drawbacks I am aware of, including its lack of usability on mobile. I am sorry to the mobile users who tried to play around with the LASR graphs; I’m sure it didn’t work well for you. Tableau is simply the best choice that I know how to use, and is good enough for now, but if anyone has suggestions or tips for a better way to display this data, please let me know.
Some potential next steps I would like to take with this project include: more advanced stats, minor league stats, foreign league stats, more granular breakdowns (i.e. LASR stats by month or week), and much more. Every night before I fall asleep I think about all the different directions I could take this. I’m not promising I’ll get there on any particular deadline, but the intention is there.
These statistics will not automatically update with new data. Right now it is the MLB offseason, so I’m safe for the next few months. However, once the 2025 season rolls around, if this project is still in its current state, I would have to manually update the data. It wouldn’t take long – probably a minute or two – but I’m a busy guy and might not always remember to do so. If you’re reading this in the middle of the 2047 season and notice it needs an update, either contact me to request an update or run back into the underground tunnel system where it’s safe.
I could write another 3000 words on this project, but I think I’ve fouled off enough pitches at this point so I’ll take my base or take my strikeout. From here, I will continue to develop LASR into its best possible version, because I care about not only my own understanding of baseball statistics but everyone else’s. I hope to establish a more informed general public about baseball stats, from the casual fan to the diehard to the 5-year old who just picked up an old Bill James Handbook for the first time and began a fateful journey into the weeds of the baseball jungle. It’s time to reimagine the stats we’ve seen our whole lives, to better understand our team’s newest addition, to equip a stronger weapon to the arsenal we use in our constant debates and WARs. It’s time to explore baseball statistics like we never have before. Enjoy!
P.S. Oh, and I guess I have to plug this new Substack of mine. Sign up!