The stats projections for the 2023 season are now up. Actually they first went up around Christmas, but I’ve been doing an unusual amount of tinkering this time around. Updates should happen at least a couple of times a week between now and the start of the season – which kicks off on March 30 this year.
1. Rules changes for 2023.
Trying to figure out how the reduced shifting and the new pickoff attempt rules will affect the statistics was definitely an exercise. A lot of it, admittedly, is guesswork, but we do have the results from the minor leagues to guide us. The gist of it is I raised stolen base attempts for all players by 25% (which obviously makes a bigger difference for guys who steal 20 bases than guys who steal 2), while decreasing caught stealing rates. Left-handed hitters received improved single rates, enough to raise their batting averages by about 10 points. Switch-hitters got a bout a 7-point raise. Righties get nothing. And similar adjustments were made for pitchers, with right-handed pitchers (who face more lefties) getting a slightly larger downgrade on their BABIP.
2. Pitcher adjustments
Every year, I make adjustments made to the way I translate between the player statistics and what I expect in the next season. The changes in the pitcher numbers are somewhat larger this year. I made a mistake last year, which did not become fully obvious to me until later, in that I way over-did the regression to mean aspect. All pitchers, good, bad, or otherwise, got squeezed into a tight bin, with not nearly enough variation. I tried to avoid that trap this year.
3. Stronger adjustments for foreign leagues
There are some oddities in projecting statistics from Japan, and I took a deep dive into them this year after a disappointing performance from some more disappointments. I’ve always calculated league strength by using the results of all players who played in those leagues. I had recognized a problem with that before – that the translations between AAA and Japan gave very different results than those between the majors and Japan – but extended it this year to noting where players are from. The result of that is Japanese hitters losing about 20 points of EqA in their translation. I will try to drw that out in another post.
4. Fielding changes
A friend of mine pointed out that the fielding numbers in the translation seemed off. This was something that I think I’ve known for a while. I had built in a sort of universal fielding adjustment to handle “defense” without having to track the position closely, and it just didn’t work. So I totally overhauled the way that worked, which should lead to more consistent values.
5. New schedules
The first set of numbers I put out still had the 2022 schedule built into it. Most years that is no big deal, but this year they are changing to a more balanced schedule, with less concentration on your own division. That was good for teams in the strong Eastern divisions (the Nationals gained 0.77 wins when I changed the schedules, most of any team; the Marlins, Red Sox, and Orioles all gained over 0.5 wins). And it was bad for the AL Central (the Guardians and White Sox were the two biggest losers, dropping by 0.85 and 0.69 wins). Those changes in team totals will be accompanied by changes in individual numbers, but it isn’t large enough to be a real issue.
6. org Files with Roster Status
One of my personal favorite tools while gearing up for the season are the org dt pages. Each team is broken down by position (with pitchers grouped into Starters, Relievers, and Swing), and each player is ranked in order of their projected Major League playing time. I find it a good way to see a team’s entire depth chart – including their coming league players. And I now have it cross-referencing with MLB roster status, so you can see who is on the 40-man roster (“A”, until we get into the season), who is an NRI (“n), and who is injured (“D”, when we get that far). Here are Brewer shortstops:
SS Last First Team Lg Age PA AB R H 2B 3B HR RBI BB SO SB CS BA OBP SLG EqBA EqOBP EqSLG EqA VORP WARP Defense FRAA MJ BRK IMP CLP ATT DRP UPS A Adames Willy MIL NL 27 597 543 81 147 28 2 28 89 53 170 7 4 0.271 0.337 0.484 0.272 0.340 0.488 0.284 43.1 5.2 140-SS 4 94 10 42 35 18 8 133 n Alvarez Eddy MIL NL 33 433 386 50 99 16 2 10 39 35 117 7 2 0.256 0.337 0.386 0.266 0.345 0.394 0.264 20.6 2.2 100-SS -1 61 16 36 56 26 8 20 A Turang Brice MIL NL 23 625 576 69 152 25 3 10 60 49 143 22 3 0.264 0.322 0.370 0.264 0.324 0.363 0.256 25.1 3.4 147-SS 5 40 27 57 30 8 1 54 n Monasterio Andruw MIL NL 26 437 398 53 97 18 2 9 43 35 116 9 2 0.244 0.311 0.367 0.258 0.326 0.375 0.249 14.1 1.3 102-SS -3 34 23 56 48 25 6 33 Devanney Cam MIL NL 26 587 539 59 117 26 2 17 59 41 165 5 2 0.217 0.281 0.367 0.231 0.294 0.376 0.233 10.1 0.9 137-SS -2 24 31 52 41 24 5 32 Brown Eric MIL NL 22 305 278 33 58 14 2 6 23 22 81 19 2 0.209 0.279 0.338 0.226 0.295 0.343 0.237 6.4 0.8 71-SS 1 15 23 40 40 1 0 50 Zamora Freddy MIL NL 24 231 218 20 46 6 1 3 13 10 63 6 1 0.211 0.255 0.289 0.211 0.254 0.284 0.202 -2.7 -0.4 54-SS -1 5 64 74 27 20 1 5 Garcia Eduardo MIL NL 20 541 516 52 110 22 2 13 53 20 217 12 1 0.213 0.250 0.339 0.227 0.263 0.350 0.214 -0.5 -0.3 127-SS -2 2 43 65 30 8 4 15 Murray Ethan MIL NL 23 427 393 38 79 16 2 8 35 31 137 9 2 0.201 0.265 0.313 0.215 0.279 0.315 0.214 -0.7 0.5 100-SS 5 2 28 51 37 13 2 11 Barrios Gregory MIL NL 19 363 341 36 78 13 2 1 28 19 82 13 1 0.229 0.275 0.287 0.233 0.279 0.289 0.214 -0.4 0.9 85-SS 8 0 61 76 22 6 1 13 Guilarte Daniel MIL NL 19 266 246 20 54 9 1 1 21 20 79 9 2 0.220 0.278 0.276 0.228 0.288 0.281 0.211 -1.0 -0.3 63-SS -2 0 49 62 27 3 0 13
I don’t know more than anyone else when games might start being played…but I do have my first look at team projections here.
Just like last year, it looks like the Dodgers and Padres are the best two teams. The Padres collapsed, epically; the Dodgers made it to the LCS before losing to the (eventual WS champ) Braves.
My spare time isn’t what it used to be…NOAA keeps me busy. But I finally do have the new MLB pbp files parsed enough to run all the split data I need. So all of the player, league, and team should be back to having those breakdowns. Not a fan of MLB’s chosen format, although to be fair I am not using their chosen tools for reading it out. It is really, really non-linear; I think my code has to make 5 separate passes through a game to get everything it needs and arranged in a game order.
Cool graph today is the playoff picture of the AL East:

The Rays (green) remain steady at the top. The Yankee (black) line surges through their August winning streak, but has been collapsing since. The Blue Jays track (blue) over the last month is a mirror image of the Yankees, with the Red Sox (red) line running right between both of them. There is a little better than a 4-in-5 chance that both AL wild cards come from the East…two out of three teams that right now are basically even.
Also in playoff odds…the Giants this morning became the first team to make the playoffs in all 1 million runs, giving them a 100% score. The Dodgers fell just short; their 99.99995 means they finished playoff tie in 1 out of a million runs, and made the playoffs in the other 999,999 runs.
The “DTs by league” pages are not updating right now, because I can’t get the play-by-play data I use to build them from MLB. Trying to figure out what’s gone wrong in my connections. The EQA report is mostly fine – except for splits data which, once again, requires the PBP files.
These come from my million-run Postseason Odds page, which will update daily through the season. Average wins by position in AL East: 100.4 91.4 84.8 77.4 58.1 AL East Pct3 Avg W Avg L Champions Wild Card Playoffs Yankees .593 97.5 64.5 63.58508 23.74102 87.32610 Blue Jays .529 86.1 75.9 13.51403 31.84269 45.35672 Rays .525 85.9 76.1 13.07292 31.31225 44.38516 Red Sox .518 84.2 77.8 9.81512 27.01932 36.83444 Orioles .370 58.4 103.6 .01285 .10404 .11689 The Yankees are an overwhelming favorite to win the division. The Orioles have the lowest playoff odds of any team in the majors. At least one of the remaining teams should take a wild card, and there's a pretty good chance that both wild cards come from the East. Average wins by position in AL Central: 94.7 86.5 79.7 72.6 63.6 AL Central Pct3 Avg W Avg L Champions Wild Card Playoffs White Sox .528 88.7 73.3 43.41552 16.64380 60.05932 Twins .527 88.1 73.9 40.12318 17.02435 57.14754 Clevelands .478 79.6 82.4 11.56375 9.88443 21.44818 Royals .444 73.7 88.3 3.96877 3.91370 7.88246 Tigers .406 67.1 94.9 .92878 .95012 1.87890 The Twins and White Sox are rated pretty close to even, with me giving the barest edge to the White Sox. The ex-Sockalexises have a chance, if their pitching can carry them through. The Royals have their best chance in several years, and some prospects whispering hopes of even better times ahead. I expect the Tigers to challenge the Orioles for the barrel bottom. Average wins by position in AL west: 96.7 85.3 78.9 73.1 65.7 AL West Pct3 Avg W Avg L Champions Wild Card Playoffs Astros .576 95.2 66.8 77.30096 7.74413 85.04509 Athletics .489 80.0 82.0 10.00943 12.37409 22.38351 Angels .477 77.4 84.6 6.30802 8.42805 14.73607 Mariners .461 75.2 86.8 4.22883 5.91044 10.13927 Rangers .441 71.9 90.1 2.15278 3.10758 5.26035 The Astros get rated as the single most likely division champion, thanks to their own strengths and contenders' weaknesses. Average wins by AL First Wild Card: 93.1 Average wins by AL Second Wild Card: 88.8 Average wins by position in NL East: 95.8 88.5 82.5 76.0 65.3 NL East Pct3 Avg W Avg L Champions Wild Card Playoffs Mets .557 88.7 73.3 36.46963 17.94029 54.40992 Braves .554 88.3 73.7 34.61201 17.93041 52.54242 Nationals .528 83.9 78.1 18.43439 14.48495 32.91933 Phillies .507 80.1 81.9 9.84990 9.46676 19.31666 Marlins .435 67.1 94.9 .63407 .78651 1.42058 A highly contentious division, with almost everyone having a chance. Average wins by position in NL Central: 93.3 86.1 80.3 73.7 61.2 NL Central Pct3 Avg W Avg L Champions Wild Card Playoffs Brewers .529 86.1 75.9 36.00577 10.24474 46.25051 Cardinals .520 84.6 77.4 29.57912 10.05231 39.63143 Reds .507 82.3 79.7 21.18838 8.59095 29.77934 Cubs .491 79.2 82.8 12.87463 6.10106 18.97569 Pirates .394 62.4 99.6 .35211 .16244 .51455 An even more contentious division, although the teams are all 1-5 games behind the East group. It is the weakest division. Average wins by position in NL West: 107.2 96.4 81.5 72.4 60.9 NL West Pct3 Avg W Avg L Champions Wild Card Playoffs Dodgers .646 104.0 58.0 63.67985 32.50205 96.18190 Padres .620 99.1 62.9 35.21897 54.38710 89.60607 Dbacks .498 77.9 84.1 .75033 11.05648 11.80681 Giants .480 74.8 87.2 .34225 6.02418 6.36643 Rockies .412 62.8 99.2 .00860 .26977 .27837 What seems very likely is that the Dodgers and Padres - with the two best projected records in the majors - will both advance to the playoffs, leaving the rest of the NL to fight over the remaining WC spot. Average wins by NL First Wild Card: 97.4 Average wins by NL Second Wild Card: 90.0
Although things don’t look much different, if we dive down under the hood you will be able to see that I have been a lot of work in the background. And I guess one of the best ways to demonstrate is to look at this page here:
http://www.claydavenport.com/ht/TORKELSON19990826A.shtml
I am now able to provide a stat-based projection on 2020’s #1 overall pick, Spencer Torkelson, based off of his college statistics. Without them I had literally nothing on Torkelson to base any stats. With them – that is a projection that makes him a very legitimate callup.
And he is not alone. Adley Rutschmann jumps from a .240 eqa projection to .284 – on the basis of his college pedigree. Andrew Vaughn likewise goes from the mid-230s to the mid-270s – because now I know he had a pair of .300 translated seasons in college.
I’m still early in my use of college stats – haven’t gotten fielding info in yet, which is going to make bring Rutschmann’s drop once it knows he’s a catcher – but it seems to work as well as minor league numbers, with a couple of caveats.
I haven’t backed these up with a formal study yet, but it certainly looks like there is a very tendency for a player’s final year in college to be a lot better than his prior years. Rutschmann and Torkelson both fit that pattern. I think that pattern will be especially pronounced for 2020 – play was interrupted typically before in-conference games started, meaning that teams (from the major conferences, at least) played unusually weak schedules. The way that I calculate the difficulties, averaging across different years, is going to mask that weakness.
The second caveat is that players appear to underperform their college numbers in their first taste of pro ball, but tend to recover towards their college averages in later seasons. There is this double whammy of a player’s final college year coming in too hot, and his first pro year too cold, that creates a LOT of distortion.
About the difficulty ratings. So the ratings I am using here see the top three college conferences as the Southeastern, the Pac-12, and the ACC. This is in line with the number of players who have gone on from these conferences to play in Organized Baseball. I counted 761 SEC players in OB ranks since 2011, 611 ACC players, and 565 from the Pac-12. Their ratings, of .498, .458, and .478 are comparable to the short-season Northwest (.469) and New York-Penn (.514) leagues (all of these numbers area average from 2015-2019). Your Midwest and Sally leagues in the .55-.57 range, the high A leagues are .63 give or take, with the majors sitting at just above 1.00 right now.
Here’s how the top 16 conferences, ranked by number of OB players, rank:
Conference Players Diff --> NY-Penn .514 Southeastern 761 0.498 Atlantic Coast 611 0.458 Pacific 12 565 0.478 --> Northwest .469 --> Pioneer .464 --> Appalachian .462 Big 12 464 0.432 Conference USA 364 0.443 Big West 347 0.433 Big 10 340 0.415 --> Gulf Coast .423 --> Arizona .418 Sun Belt 274 0.407 West Coast 263 0.420 Southland 258 0.388 Big East 266 0.432 Western Athletic 242 0.372 Missouri Valley 219 0.432 Mountain West 217 0.435 American Athletic 214 0.413 Atlantic Sun 205 0.411
There is a .81 correlation between the number of players taken from each league and its quality, so the draft appears to be resosnably efficient The ratings do tend to keep dropping from there. By the time you get to, say, the Wisconsin Intercollegiate League, made up of the smaller schools of the University of Wisconsin system (like Oshkosh, Stevens Point, or Whitewater), the rating is coming in at just .130.
Hat tip to baseballreference.com, who really pulled these collegiate numbers together. My task was just to pull them together in a sensible fashion
Major League Baseball has never had a player whose last name starts with X. Joe Xavier, who made it to AAA in 1989-90, is as close as any X player has gotten.
Until now.
Major League Baseball’s decision to reclassify some of the Negro Leagues as “major league” changes that. The Negro American League – once of the reclassified leagues – had a man by the name of Leovigildo Xiques, who played for the (Cincinnati and) Indianapolis Clowns.
The title says it all. They are under “DTs by League“.
First things first – I now have the way too soon, already out of date, Team Projections for 2021 up and available.
Of course, posting them on the day of the non-tender deadline means most teams have at least one player in my projection that is already compromised. And of course, teams have yet to fill the gaping holes (like the Nationals 1B slot) that you know will be taken by some current free agent. Still, its nice to have a snapshot of where the team stood when they started making decisions about 2021.
generated on 12- 2-2020, projections for 2021 season
AL East Won Lost Runs RunsA NYY 92 70 845 737 TBY 86 76 769 726 BOS 83 79 837 816 TOR 78 84 779 803 BAL 70 92 707 808 AL Cent Won Lost Runs RunsA MIN 87 75 809 747 CLE 84 78 764 749 CWS 82 80 788 772 KCR 72 90 714 799 DET 69 93 673 785 AL West Won Lost Runs RunsA HOU 89 73 818 742 LAA 83 79 812 792 OAK 81 81 743 743 TEX 77 85 755 783 SEA 71 91 703 791 | NL East Won Lost Runs RunsA ATL 91 71 815 719 NYM 86 76 752 707 WAS 83 79 735 723 PHI 80 82 738 750 MIA 74 88 673 739 NL Cent Won Lost Runs RunsA MIL 86 76 750 706 CHC 83 79 756 738 CIN 82 80 755 748 STL 77 85 707 744 PIT 73 89 700 777 NL West Won Lost Runs RunsA LAD 96 66 843 704 SDP 89 73 779 707 ARI 81 81 768 768 COL 77 85 783 819 SFG 67 95 665 792 |
One thing I am pretty certain about – not just for me, but everybody else doing projections, from BP to FanGraphs to ESPN, Rotowire, Rotoworld, and every other fantasy site – projections are going to be worse in 2021 than in prior years. Having a 2020 season that was only 60 games – and even then, only for the majors – every player is coming off of what we would, previously, have labelled a missed or truncated season. Our sample sizes are smaller than ever.
To balance that, I tried something I never did before. The “2020” season that is going through the program, combining with 2019 and 2018 stats to make a 2021 projections, is a blend of whatever the player really did in 2020 with the projection I made going in to 2020. For every minor leaguer, who had no 2020 season at all (or at least nothing with public statistics!), there is nothing to blend – the 2020 projection is their entire line.
I can’t say I’m enthused with the results, but testing says it was a lot better than going with zero seasons or taking 2020 results at full face value.
Welcome to anyone who might be checking in on me for the first time. Or for the first time in a long time. Either way is good for me.
As you can plainly see, I don’t write that often. At heart, I am a number-cruncher, not a wordsmith, and this site is about the numbers that are behind the tabs, not the words I use to highlight or stretch those numbers. And that in turn is because I am not thinking of this as a public site; it is really a vessel for me to have a lot of this work available, to me, when I am not sitting at my own desk.
So a quick guide to whats on here…
The EQA Report tab is for the major leagues, only. I have tables for all major league seasons, back to 1871, which update every day in season. The centerpiece of these tables is a statistic I called Equivalent Average, which measures a players total offensive value and scales it to the historical range of batting average. Not because batting average is superior in any way – it certainly is not – but because baseball fans, at least of my generation, have an immediate grasp of the scale.
Next up is the Playoff Chances tab. This is the output of a million-run Monte Carlo simulation of the rest of the season – at this point, the entire season. My version has a wider spread than similar charts run by other sites; I believe they are too settled on their valuation of (say, the Yankees) as a “.600 team”. I use some tricks to build a distribution of values for the Yankees. In most runs, they are a .600. But in some runs they’re a .650 team; in others, and more, because it is easier for things to go wrong than right, they only get treated as a .550 team. That’s how the chances for even the Orioles, while miniscule, are not zero.
Then is DTs By League. These are statistics for minor league players, organized by league. There are three flavors to these pages. The first flavor is the real stats – completely unadjusted. The second flavor is the DT – the Davenport Translation, originally developed in the late 80s – which estimates what that the real statistical line is worth, right now, in major league terms. The third flavor – peak DT – builds on the second flavor but adds “normal” improvement with age. For the US minor leagues, I have these stats going back to 1979. Triple-A leagues go back to 1946. Japanese leagues go into the 1930s. Other foreign, winter, and independent leagues are randomly available.
DTS by Organization are probably the pages I myself look at most often during the season. It is only going to have players for the current season, arranged by major league team, with all of their minor league affiliates listed with them.
The Projected Standings is where projected statistics, for team and players, resides. The projections have two flavors to them. One is the straight out of the computer numbers; the other is the result of loading those numbers into a very large spreadsheet, manually assigning playing time to everyone, and then rebalancing the numbers so that the hitting and pitching totals are consistent. Those manual adjustments are a real chore, and difficult for me to maintain through the season.
The last tab is the Spring Stats, which is going to start populating Any Day Now. While they are generally of limited predictive value, they certainly do affect who actually makes it out of spring training. There is also a second tab in there, for Current Team DTs, which has three entries per team – all of the team’s hitters (“.2020dt”), all of their pitchers (“.2020pdt”), and a depth chart (“.2020orgdt”) culled from the computer-only projections.
While everything on this site is free, a donation through Paypal to help offset costs would be greatly appreciated. -Clay
If you are trying to reach me, drop me an email. Same address as the webpage, but replace ".com" with "@gmail.com".
Archives
- January 2023
- January 2022
- September 2021
- April 2021
- February 2021
- December 2020
- February 2020
- November 2019
- January 2019
- March 2018
- February 2018
- January 2018
- August 2017
- June 2017
- March 2017
- January 2017
- September 2016
- May 2016
- April 2016
- March 2016
- February 2016
- January 2016
- September 2015
- April 2015
- March 2015
- January 2015
- December 2014
- November 2014
- October 2014
- April 2014
- February 2014
- January 2014
- October 2013
- April 2013
- March 2013
- February 2013
- March 2012
- February 2012
- January 2012
- December 2011
- November 2011
- October 2011
- August 2011
- July 2011
- June 2011
- May 2011